Site Selection Characteristics of Shrines Affected by Tsunamis in the Coastal Areas of the Sanriku Region

Journal of JSCE, Vol. 5, 226-245, 2017 Special Topic - Restoration and Recovery from the 2011 Great East Japan Earthquake（ Invited Paper）

SITE SELECTION CHARACTERISTICS OF SHRINES AFFECTED BY TSUNAMIS IN THE COASTAL AREAS OF THE SANRIKU REGION

Shin OSAKI1

1Member of JSCE, Assistant Professor, Dept. of Civil Eng., University of Tokyo (Hongo 7-3-1, Bunkyo-ku, Tokyo 113-8656, Japan) E-mail: [email protected]

This study aims to define site selection characteristics of shrines affected by tsunamis in the coastal ar- eas of the Sanriku region, to understand the differences between the locations of shrines that did or did not survive the disaster through field surveys and historical reviews. First, through the field survey, this research demonstrates that shrines in the area are generally located on ridge of mountains less likely to be affected by tsunamis and the destroyed shrines tend to be located near the low-lying coastline or on flat- lands. Second, through the historical reviews, the characteristics of shrine locations have greatly changed since the middle of the 12th century. This research indicated that shrines were mostly positioned along the ridge line during the time of the ancient belief system and became more diverse during the medieval period from the mid-12th century onwards, when the culture of “kanjo,” the ceremonial transfer of a di- vided tutelary deity to a new location according to the enshrined deity’s character, was introduced to the area. In addition to this general site selection characteristics, the experience from earlier tsunamis affects the re-selection of shrine sites. In conclusion, the site selection system for the shrines consists of global and local mechanisms. Finally, the author suggests the possibility of utilizing the shrine’s site selection characteristics as a way to hand down the tsunami experience to the next generations.

Key Words : shrine’s site selection, shrine founding, transferring deity, Sanriku Region, tsunami

1. INTRODUCTION moved back to where they once lived 2). Yamaguchi explains that people’s nostalgia for their homeland (1) Background and the convenience of work have acted synergisti- Cities and villages develop unique forms over a cally to make people want to resettle in their origi- long period as people enjoy the blessings of nature nal locations. while making necessary adjustments to protect The low frequency of disasters caused by tsuna- themselves from its power. The villages settled mis seems to be closely associated with people’s along the coastal areas of Sanriku have prospered preferences to live in regions on lower land. mainly through the fishery industry, despite being Large-scale tsunamis often occur beyond the life battered repeatedly by tsunamis. In the Sanriku re- span of most individuals, thus, exceeding the max- gion, where extensive damage was caused by the imum period that the experience of a tsunami can be tsunami that occurred alongside the 1933 Sanriku handed down directly from one generation to the earthquake, the Secretariat of the Minister of Home next. In recent years, people have tended to overes- Affairs’ city planning section implemented a mod- timate safety, believing that sea walls built along- ernized recovery plan. Accordingly, residents of side the development of civil engineering technolo- fishing and agricultural villages, which are easier to gy will protect their land from the threat of a tsuna- relocate, have collectively been moved to higher mi. For these reasons, people from the Sanriku ground. Meanwhile, residents of cities with popula- coastal area may have opted for the convenience of tions too large to relocate have been directed to re- work and decided to settle on lower land, rather than settle in the locations where they once lived1). protecting themselves from the threat of another However, Yaichiro Yamaguchi reported that in the tsunami. On the other hand, shrines have existed in years since then, people from some of these villages or around the village for much longer, far beyond

226 the scope of people’s memories. place in the Japanese shrine origin has totally dif- Shrines have been worshiped by the people of the ferent geographical and historical features from the village as symbols of their beliefs. At the same time, coastal areas in Sanriku region so that the site selec- they have functioned as the village’s key public tion mechanism of shrines in this research site is space. The Sanriku coastal area is also known as a still unclear. region where local residents enthusiastically cele- One of the most important attributes is the en- brate religious ceremonies at shrines. Traditional shrined deity. Enshrined deities are known to have performing arts such as Toramai (tiger dance) and unique characters. In general, these are often recog- Shishi-Odori (deer dance) are handed down with nized as “divine virtues” or “divine favors.” local resident’s proactive support. Thus, as many Well-known examples include Inari, who is wor- shrines in the Sanriku coastal area often function as shiped as a deity of agriculture and vocations in key community hubs, they can be considered as in- general, Michizane Sugawara as deity of learning, dispensable key locations when we conduct spatial and Ebisu as deity of business prosperity. Urasaki8) planning for recovery and regeneration of the com- conducted a research on the locations of Mikumari munity. shrine (“mikumari” stands for “dividing water” and As described below, much news coverage and the shrine is known for its deity of water distribu- many previous studies emphasized that many tion) and Takagama shrine, Nara, which is known as shrines barely survived the tsunami in 2011. How- a shrine to pray for rains, and pointed out certain ever, it is important to note that not all shrines sur- relevance between the enshrined deities and vived. There is a need to know the differences be- topographic characteristics. tween the locations of shrines that did or did not According to these researches, the topographical survive the disaster. locations of the shrines tended to differ from region to region, whereas the tendency seemed to be af- (2) Purpose fected by the shrines’ attribution such as sizes and Based on this background, this study aims to de- enshrined deities. Enshrined deities should be cho- fine site selection characteristics of shrines affected sen under strong influences of local conditions such by tsunamis in the coastal areas of the Sanriku re- as climates, landshapes, occupation of the citizens, gion, to understand the differences between the lo- and disasters, for example. That is, enshrined deities cations of shrines that did or did not survive the dis- are supposed to have an influence on the shrines’ aster. Furthermore, it discusses how the site selec- site selection from a topographic point of view. tion characteristics of shrines can be utilized for re- generation. (2) Shrines’ location characteristics compared with disasters According to the previous researches on the asso- 2. PREVIOUS RESEARCHES ciations between shrine locations and natural disas- ters, focused on the area surrounding Hanaore Fault, (1) Shrines’ location characteristics from a Kyoto (Koresawa and Horikoshi, 20049)) and Ko- topographic point of view surido area, Toyoma (Hattori and Nii, 201210)), Early researches on the characteristics of shrine shrines tended to stand in the locations which were locations typified from a topographic point of view likely to be affected by natural disasters or in the focused on Fukuoka prefecture (Hitaka et al., locations meant to protect the village from natural 20053)) and pointed out that there were many shrines disasters. situated in flatlands and basins. Other researches On the contrary, in the coastal areas of Tohoku that focused on Saga plain (Goto and Nakaoka, region, which are quite frequently affected by tsu- 20024)), Nara prefecture (Urasaki, 20085)), and nami, people often speak of shrines as locations that Tohoku region (Tagami, 20046)) reported that the are less likely to be affected by natural disasters. A shrines in these areas tended to be located at the feet good example is Namiwake shrine in Wakabayashi, or top of mountains. In other words, the characteris- Sendai, Miyagi. Imamura introduces an old myth tics of shrine locations varied according to the re- related to this shrine, which is believed to have di- gions. vided a gigantic tsunami into two and saved the res- Further, Urasaki7) reported that the geographical idents from damage (Imamura, 201111)). Further, features of the shrine location varied in accordance Miyasaka et al.12) researched on the Sennan plain, with the size of the shrine, which was a useful result where Namiwake shrine is located, and found that of shrine location analysis where attributes of shrines tended to be located on slightly higher shrines were taken into account. Whereas his re- grounds such as beach-ridges and natural levees in search site, Nara prefecture, one of the most historic those villages located closer to the sea, even in a

227 vast flatland area. Further, Takada et al.13) conducted be influential on the shrines’ site selection. Howev- an intensive research on the damage situation of er, there are still many shrines that do not belong to shrines located along the coastal areas of Miyagi the Association of Shinto Shrines, and the shrines caused by the tsunami in 2011, with analysis focus- whose origins are unknown are not necessarily ing on the enshrined deities. This research indicated proven to have long histories. There are also no re- that there could be some relevance between the search results with factual evidence that identifies number of damaged shrines and the characters of the the association with the previous tsunamis. enshrined deities; for instance, the shrines where Accorindg to the research conducted by Endo and Susanoo-no-Mikoto, the deity of water and natural Mazereeuw18) that focused on Shizugawa, Utatsu, disaster was located were much less likely to have and Tokura areas of Minami-sanriku, Miyagi, older suffered damages. shrines built in former castle sites in the Middle According to these researches, shrines in the Ages were less likely to be affected by disasters; coastal areas of Tohoku region tended to be located whereas those newer shrines, which were built on in a topography safer from tsunamis. In addition, lower grounds closer to the downtown area in or enshrined deities had associations with disasters and after the Edo period when the political situation was showed significant relevance with damage from the stable, were more likely to suffer damage. This is tsunami in 2011. In other words, there were two the latest piece of knowledge concerning the factors that affected shrine’s location from a shrines’ location characteristics associated with their topographic point of view: tsunamis and enshrined historical and local backgrounds. deities, and these factors are associated with each other. (4) Originality of this research As briefly described in the previous sections, (3) Shrines’ location characteristics compared shrine locations vary by regions in Japan and the with disasters and history earlier researches showed that such variety was Takase et al.14) conducted a field survey of the caused by several factors: attributes of shrine (size damage status of shrines in the coastal areas in Fu- and enshrined deity), natural disasters (tsunamis), kushima prefecture and reported that shrines tended and local history. However, in the coastal areas of to be located around the borderline of tsuna- Sanriku region, no previous wide-area research has mi-inundated areas. This research also showed that demonstrated a relationship between the locations most of the shrines that survived the tsunami had and damage to shrines, including the ones that are unknown provenance, while many of those damaged not members of the Association of Shinto Shrines. shrines had written records of their origins. This Further, although there are individual researches that implied that newer shrines with records of historical point out the importance of the shrine’s history, at- trails had been damaged. There were a number of tributes, and past tsunamis relevant to the shrine’s related discourses15) including a reported quote of site selection characteristics, no previous research the Association of Shinto Shrines: “shrines with has done a comprehensive analysis of all these longer histories have survived the tsunami associat- points based upon historical records. Thus, the orig- ed with the Great East Japan Earthquake, and many inality of this research lies in the comprehensive of them stood on the borderline of tsuna- objects and analysis of the shrine’s site selection mi-inundated areas.”16) In addition, Matsuura17) re- characteristics in the area. searched the damage status of the member shrines of This research mainly focuses on the history of the Association of Shinto Shrines Iwate and Miyagi shrines and previous tsunamis to analyze the prefectures and reported that the shrines were lo- shrines’ location characteristics and their damage cated in the areas that were free of tsunami damage situation from the tsunami in 2011 based upon his- in Iwate, whereas they were not damage-free in torical records. Further, this research deals with the Miyagi. The research reported that the difference combination of the shrines’ history with their attrib- derived supposedly from the fact that the tsunamis utes. that hit Miyagi prefecture in and after the Meiji period were lower in height compared to those in Iwate prefecture and that there were hardly any 3. OBJECTS AND METHOD memories that remained of the tsunami associated with the 1611 Keicho Sanriku earthquake because (1) Site most of the villages in Sendai plain had once turned The research site was the coastal areas of the into wastelands due to the tsunami of 1611 but were Sanriku region from Kuji Bay in Kuji City to cultivated after that. Mangokuura Lake in Ishinomaki City (islands are Thus, the historical backgrounds of the site might not included). In these areas, the damage from the

228 tsunami in 2011 was especially severe (Fig.1). In more shrines identified through field surveys were total, 15 municipalities were included in the re- added, to total 466 extracted shrines. search site as follows: - Iwate prefecture : (3) Survey method Kuji City, Noda Town, Fudai Village, Tano- First, we arranged information on the location of hata Village, Iwaizumi Town, Miyako City, extracted shrines on one GIS map. Then, we deter- Yamada Town, Otsuchi Town, Kamaishi City, mined the damage the tsunami in 2011 caused to the Ofunato City and Rikuzen-Takata City shrines through 12 field surveys conducted from - Miyagi Prefecture : November 2011 to September 2013 by our survey Kesennuma City, Minami-Sanriku Town, On- team composed of the author and Yuta KANAI with agawa Town and Ishinomaki City assistants from the author’s laboratory. Second, we Although it was not easy to define the exact realm collected data on the attributes and histories of the of the coastal area, this research focused on areas extracted shrines by examining local history books, affected by the tsunami; thus, the area must be de- the Shrine Directories of Iwate and Miyagi Prefec- fined in association with the inundation area of the tures, other books related to the shrines (Ex: Mit- tsunami in 2011. Therefore, the coastal area was suhashi and Shirayama, 200521)), historical explana- defined as including all villages in or next to the tion boards on shrines, and interviews conducted inundation area. with local residents.

(2) Shrine extraction method (4) Composition of this research The shrine extraction method was applied to in- First, Chapter 4 reports on the damage to shrines clude as many shrines in the area as possible, in- caused by the tsunami in 2011 and confirms the cluding those that do not belong to the Association relevance to attributes (size and enshrined deities) of Shinto Shrines. and damage. In Chapter 5, based on historical re- First, we conducted a keyword search of GIS data search, the author highlights general site selection of Zmap-Town II 2008/09 Shape version (Zenrin characteristics for the shrines. Chapter 6 examines Co., Ltd.), and then extracted those that seemed to the local histories of shrines inundated in previous be the names of shrines located in the coastal area. tsunamis. Finally, the author combines the general This process resulted in extracting 366 shrines. Se- and local findings explained above to provide con- cond, we compared the GIS map with topographic clusions regarding site selection characteristics for maps of 1:25,000 (Geographical Information Au- shrines influenced by tsunamis and to discuss re- thority of Japan) and added another 82 shrines. generation. Third, we compared the extracted shrines to the Shrine Directories of Iwate19) and Miyagi20) Prefec- ture and added another 15 shrines. Finally, three 4. DAMAGE TO SHRINES CAUSED BY THE TSUNAMI IN 2011

(1) Classification of shrines based on damage Our field surveys could identify 443 of the ex- tracted 466 shrines. We could not identify the re- maining 23 shrines for various reasons: they were not found at the place indicated on the map, or the approach path was designated as a non-passing zone by a landslide, etc. We classified shrines into three categories according to damage sustained to their elements, namely the shrine (building), Sando (ap- proach road(s) to a shrine), and Torii (gateway(s) built at the entrance and/or midway of Sando) as follows (Table 1): - Destroyed: The shrine was destroyed or washed away. - Damaged: The shrine was partially damaged, or The Sando and/or Torii were damaged. - Undamaged: Fig.1 Research site. No damage was evident for any of the three

229 Table 1 Images of classification of shrines based on damage. Destroyed Damaged Undamaged The Shrine was partially damaged, or No damage was evident for any of the The Shrine was destroyed or washed away The Sando and/or Torii was damaged three elements

Niwatari Shrine Dewa Shrine Tateishi Shrine On Takashirahama Area in Onagawa Town On Taro area in Miyako City On Ryouri area in Ohfunato City

elements. Table 2 Numbers of damaged/undamaged shrines located As there was no research on the damage to in/out of inundation area of the tsunami of 2011. shrines in Sanriku region at the moment we started Un- Destroyed Damaged SUM the survey, this classification is our original one. damaged Afterwards, researches could be found on the shrine Shrines in inun- 54 27 22 103 damage as follows: Takada et al.22) surveyed shrines dation area (52%) (26%) (21%) (100%) that could be found on maps and classified them Shrines out of 6 77 257 340 inundation area (2%) (23%) (76%) (100%) into three categories of undamaged, partially dam- Total 60 104 279 443 aged, and damaged although there was no descrip- tion of the criteria. There were 33 shrines in their research that we also surveyed. Two of them were (2) General tendency of damage to shrines classified as “damaged” in their research but were a) Result of totaling data from the field survey classified as “undamaged” in this research. These In total, 60 shrines (14%) of 443 damaged during examples indicate possibly that our survey team the tsunami were classified as “destroyed,” 104 overlooked the clue of damage. However, the num- (23%) as “damaged,” and the remaining 279 (63%) ber of misjudgments is very limited so that the gen- as “undamaged.” The ratio of “destroyed” shrines is 23) eral tendency seems to be unaffected. Matsuura similar in both Iwate and Miyagi Prefectures. reported the damage to shrines based on the records b) Shrines located in the inundation area that of the Association of Shinto Shrines that was a col- were not damaged lection of reports from each shrine. There were sev- Buildings located in the inundation area were eral shrines classified as “undamaged” in his re- usually severely damaged unless the area had fea- search but were classified as “damaged” in this re- tures that minimized the power of wave surges, such search and vice versa. That seems to have happened as geographical features. Table 2 shows the number because of the character of the records. Shrines of shrines not damaged despite being located in the might not have reported to the Association of Shinto inundation area. In total, 49 shrines (48%) of 103 Shrines incidence of light damage such as scratches sustained no damage to the shrine (building), mean- on Torii (NO2, Atago Shrine, etc.). But the fact that ing that shrines in the area tended to be located the tsunami reached the Torii was important for our away from the impact of the tsunami’s wave surges. research. On the other hand, shrines might have re- As shown in Table 2, six shrines which were not ported the missing properties that were not kept in in the inundation area were classified as “de- the shrines but elsewhere, like the residence of the stroyed.” It means either the author’s shrine damage priest. For example, Minato Ohsugi Shrine (MI9) classification or the inundation area data from reported that “the implements for the traditional fes- Haraguchi25) were wrong. For example, the author tival were all washed away” in the report, but the might have misjudged the cause of damage to the shrine stood at an altitude of over 50m and that was shrine building; the author might have thought it higher than the highest records of the tsunami reach. was derived from the tsunami although the building 24) Takase et al. also did field surveys mainly in the might have actually been damaged during the Fukushima prefecture. We also surveyed two earthquake. Otherwise, Haraguchi’s survey might shrines and found the same results in both shrines. have misidentified the exact inundation area. How- Thus, it can be said that the author‘s field surveys ever, six is an allowable margin of error for the con- have proper results in general. clusion that the main buildings for half of the

230

Altitude (m) LEGEND 60.0

50.0

40.0

30.0

20.0

10.0

0.0 0 1000 2000 3000 Horizontal Distance from the Nearest Coastline (m) Fig.2 Relationship between the shrine’s position (distance and height from the nearest coastline) and damage sustained. shrines located in the inundation area were undam- aged. c) General tendency of damage to shrines The tsunami in 2011 completely destroyed or washed away 60 shrines (14%). However, the main buildings for half of the shrines located in the inun- dation area were undamaged. The results indicated that shrines in the Sanriku coastal area were gener- ally located in areas less likely to be affected by tsunamis.

(3) Relationship between the shrine’s location and damage sustained a) Relationship between the shrine’s position Fig.3 Classification of shrines of topographical features. (distance and height from the nearest coast- line) and damage sustained The author extracted the altitude of the shrines Shrine located on the side of a mountain ridge. from 10m mesh DEM data (Geographical Informa- - V Type : tion Authority of Japan) and calculated the horizon- Shrine located deep within a valley. tal distance from the nearest coastline using coast- - F Type : line GIS data from the National Land Numerical Shrine located on a flatland. (These shrines are Information Download Service. located only in a residential area.) The horizontal axis in Fig.2 represents the dis- - C Type : tance from the nearest coastline, while the vertical Shrine located near the coastline. (These axis represents the altitude of the shrine. Figure 2 shrines are located out of a residential area.) illustrates that the destroyed shrines were generally - M Type : located near the low-lying coastline. The graph in- Shrine located on a highest peak of a mountain. cludes 432 shrines that were located within 60m - I Type : altitude and 3000m distance. Shrine located on an isolated hill on a flatland. b) Relationship between the topographical fea- As seen in Table 3, the most popular locations tures and damage sustained are T- and S-type shrines: 311 (72%) of 443. In The author classified shrines into seven types of total, 357 (82%) of the 443 shrines were classified topographical features as follows (Fig.3): as having mountain-related topographical features - T Type : (T, S, M, and I Types). M-type shrines are those Shrine located at the top of a mountain ridge. with the safest locations, as none were “destroyed.” - S Type : The most damaged shrines were those in the C- and

231

Table 3 Numbers of damaged/undamaged shrines sorted by classifications of topographical feature of shrines. Destroyed Damaged Un- SUM damaged 12 57 111 180 T Type (7%) (32%) (62%) (100%) 10 25 96 131 S Type (8%) (19%) (73%) (100%) 1 4 12 17 V Type (6%) (24%) (71%) (100%) 27 5 20 52 F Type (52%) (10%) (38%) (100%) 9 4 4 17 C Type (53%) (24%) (24%) (100%) 0 1 17 18 M Type (0%) (6%) (94%) (100%) 1 8 19 28 I Type (4%) (29%) (68%) (100%)

F-type categories, wherein more than 50% were “destroyed.” c) Relationship among the shrine’s position, topographical features, and damage As seen in Fig.4, which is a scattered plot of shrines within 60m altitude and 2,000m horizontal distance from the nearest coastline, F type includes many of the lower plots and C type includes lower and nearer plots than others. It is also unique that F type contains “destroyed” shrines, which are not necessarily near the coastline. That is, combined with the findings on section a), the “destroyed” shrines tend to be located near the coastline in low height or on flatlands. Thus, it is confirmed that topographical features are related to physical posi- tions (altitude and distance from the sea) and those factors are related to damage from the tsunami in 2011 as well.

(4) Relationship between the shrine’s attributes and damage sustained a) Relationship among the shrine’s size, location, and damage sustained The author classified shrine size into four cate- gories according to the presence of all or some of the following three elements: House of worship (Haiden), Torii, and Yard. This classification is based upon Urasaki’s26). Table 4 and Table 5 show the number and ratio of shrines. - Large: Shrines comprising all three elements, namely a House of worship (Haiden), Torii, and Yard. - Medium: Shrines comprising two of the three elements. - Small: Shrines comprising one of the three elements. - Tiny: Shrines comprising no elements of the three Fig.4 Relation between position and damage in a topographical listed. (The shrine is a small box and not a features classification. building.)

232

The ratio of “destroyed” shrines (47%) is espe- which tended to be located in the flatlands that were cially high for “tiny” shrines. However, no major exposed to a risk of the tsunami, no significant dif- differences were found for the other sizes. ference was found among shrine’s size, location and Seventeen shrines (45%) of these 38, more than damage. for any other classification, were located in the flat- b) Relationship among the enshrined deity, loca- lands (F Type). In addition, out of 18 “destroyed” tion and damage sustained “tiny” shrines, 14 shrines (78%) were located in the Table 6 provides the research results for the pri- flatlands. This implies that the “tiny” shrines were mary enshrined deities, damage sustained, and the chosen to be built when people wanted to build a shrines’ topographical classifications. Figure 5 il- shrine in their residential areas, because the flatlands lustrates the scattered plots of shrines within 60m were very limited in the areas of the Sanriku region. altitude and 2000m horizontal distance from the Supposedly, most “tiny” shrines are not for the use nearest coastline sorted by enshrined deities. of the community, but for private individuals who “Susanoo” descent is a shrine whose primary en- build it at their residences. Thus, the site selection shrined deity is Susanoo-no-Mikoto, who is a deity process for these types of shrines might be different. of flood and plague that occurs after the flood. In conclusion, other than the “tiny” shrines, “Kumano” descent is a shrine whose name has “Kumano” and it has three enshrined deities. One Table 4 Numbers of topographical features category sorted by of them, “Ketsumimiko-no-Ookami” is identified size of shrines. as Susanoo-no-Mikoto. Thus “Kumano” descent T S V F C M I SUM should have close characteristic as “Susanoo” de- 56 36 6 8 2 5 10 123 Large scent. “Hachiman” descent is a shrine whose pri- (46%) (29%) (5%) (7%) (2%) (4%) (8%) (100%) Me- 70 47 2 15 5 2 11 152 mary enshrined deity is “Hondawake-no-Mikoto” dium (46%) (31%) (1%) (10%) (3%) (1%) (7%) (100%) who is identified as a Oujin emperor. The emperor 45 32 7 10 6 7 2 109 Small is known as a person who introduced technology (41%) (29%) (6%) (9%) (6%) (6%) (2%) (100%) (mainly for infrastructures) and culture into Japan 6 7 2 17 2 3 1 38 Tiny (16%) (18%) (5%) (45%) (5%) (8%) (3%) (100%) from the continent. “Inari” descent’s primary en- Un- 3 9 0 2 2 1 4 21 shrined deity is “Uka-no-Mitama,” deity of agri- known (14%) (43%) (0%) (10%) (10%) (5%) (19%) (100%) culture and vocations in general. “Amaterasu” de- scent ‘s primary enshrined deity is “Ama- Table 5 Numbers of damaged/undamaged shrines sorted by size terasu-Oomikami,” the sun goddess. “Kompira” of shrines. descent is a shrine whose primary deity is “Oo- Un- Destroyed Damaged SUM mono-Nushi-no-Kami,” the deity of voyage, fish- damaged 9 34 80 123 ery, and water. “Munakata” descent’s primary dei- Large (7%) (28%) (65%) (100%) ties are three goddesses who represent soothing the 15 47 90 152 Medium nation and voyage. (10%) (31%) (59%) (100%) Takada et al.27) reported that although their site is 8 23 78 109 Small (7%) (21%) (72%) (100%) in the coastal areas of Miyagi prefecture, Ofunato 18 0 20 38 City, Rikuzen-Takata City in Iwate Prefecture, few Tiny (47%) (0%) (53%) (100%) shrines were destroyed for the deities of Susanoo, Un- 10 0 11 21 Kumano, and Hachiman descent, although many known (48%) (0%) (52%) (100%)

Table 6 Number of damaged/undamaged shrines and topographical classification sorted by enshrined deities. Damage Classification Topografical Classification Descent Name Un- SUM Destroyed Damaged T S V F C M I damaged 2 5 17 12 8 0 1 0 1 2 24 Susanoo Descent (8%) (21%) (71%) (50%) (33%) (0%) (4%) (0%) (4%) (8%) (100%) 2 4 18 6 9 2 4 0 0 3 24 Kumano Descent (8%) (17%) (75%) (25%) (38%) (8%) (17%) (0%) (0%) (13%) (100%) 1 9 20 13 7 2 1 2 3 2 30 Hachiman Descent (3%) (30%) (67%) (43%) (23%) (7%) (3%) (7%) (10%) (7%) (100%) 6 11 30 21 12 5 7 0 1 1 47 Inari Descent (13%) (23%) (64%) (45%) (26%) (11%) (15%) (0%) (2%) (2%) (100%) 2 12 20 16 14 1 0 0 1 2 34 Amaterasu Descent (6%) (35%) (59%) (47%) (41%) (3%) (0%) (0%) (3%) (6%) (100%) 1 5 11 7 4 0 1 1 1 3 17 Kompira Descent (6%) (29%) (65%) (41%) (24%) (0%) (6%) (6%) (6%) (18%) (100%) 1 5 4 3 2 0 1 3 0 1 10 Munakata Descent (10%) (50%) (40%) (30%) (20%) (0%) (10%) (30%) (0%) (10%) (100%)

233 were destroyed for the deities of Inari and Ama- terasu descents. General tendencies except for the Amaterasu de- scent identified in this research related to the en- shrined deity and damage to the shrine are similar to their research, although that is not as vivid as theirs. The ratio of “destroyed” shrines (13%) is rather high for Inari descent shrines, which has higher ra- tio in those located in the flatlands compared to other descents. This is because the Inari descent, a deity of agriculture and vocation in general, has re- lationship with flatlands and/or residential areas. Figure 5 shows the diversity of Inari descent’s loca- tions, whereas Amaterasu descent, which is located mainly by the coastline in 10m-30m altitude, has a certain character in their location. This tendency can be seen on the Kompira descent, one of the leaset damaged shrine in this research, and Susanoo de- scent although it is about 10m higher than the Ama- terasu descent. Susanoo, a deity of flood, and Kom- pira, a deity of voyage both have reason to be lo- cated on a higher place near the coastline to watch the safety of port and/or ships on the sea. For the same reason, Munakata descent, a deity of voyage, is located very close to the coastline. However, the reason for Amaterasu descent’s location could not be identified in this research. In conclusion, some deities have their characters in their location according to their characters as dei- ties. In addition, the location tendency is related with the damage sustained. c) Summary First, the author demonstrated that shrines in the area were generally located on ridge of mountains less likely to be affected by tsunamis and the de- stroyed shrines tended to be located near the low-lying coastline or on flatlands. Second, the investigation on the size attributes of shrines indicated a minimal relationship with dam- age except for tiny shrines, which were not build- ings, just boxes, and tended to be located in flat- lands. However, the investigation on the enshrined deity indicated relationship with location and damage. In some deities, Susanoo, Inari, Kompira, Munakata descents of shrines for example, the location ten- dency could be explained by those deities’ charac- ters.

5. GENERAL CHARACTERISTICS OF SITE SELECTION FOR SHRINES

(1) Time when the shrine was founded This section presents the results of the author’s Fig.5 Relation between position and damage sorted by research on when each shrine was founded and other enshrined deities.

234 background information. (Emperor Ojin) or Bishamonten, both related to the The author found written records on when 119 of armed forces, corresponds to the hypothesis that the all 466 shrines investigated in this research were shrines were founded during the war. Furthermore, founded. The time of founding and accumulated written records stipulate that Ama-terasu-mioya number of shrines are plotted in Fig.6. If the record Shrine was founded for people to pray for the repose reads “founded in the Daido era” for example (the of the souls of dead enemies, providing further evi- Daido era started in 806 and ended in 809), then the dence that the shrines were founded directly after last year of the era is plotted on this graph. (In the the troops departed. If the shrines were founded earlier example, the founding year is listed as 809.) during periods of war, they could have theoretically The graph indicates a peak period when several been founded during any historical era. However, shrines were founded simultaneously. These are hardly any written records exist as evidence. Written numbered as periods I through V. There was no no- records on the founding of the aforementioned table relevance between the peak period and histor- shrines were locatable because the wars were caused ical tsunami events, which disproves the theory that by Imperial troops, the authorized history of which more shrines were founded because of the occur- was recorded. rence of a tsunami. (3) Period II (around 1050): founding of shrines (2) Period I (around 800): founding of shrines and background and background Period II includes six shrines founded during the Period I includes six shrines founded in the period period 1040–1064. The research results from docu- 801–809. The research results from documents re- ments related to the history of the founding of lated to the history of the founding of shrines are shrines are provided in Table 7. As with Period I, provided in Table 7. Records indicate that an Impe- many shrines were founded because of wars. Writ- rial troop General Sakanoue-no-Tamuramaro, was ten records indicate that the so-called Former Nine involved in founding all shrines—except Haguro Years’ War (1051–1060) waged by Yoriie and Yo- Shrine—when he went on an expedition to the east, shiie of Minamoto against the Abe clan led to the which explains the synchronicity in terms of timing. founding of four shrines. Further, these four shrines Specifically, these shrines were founded when Sa- were all located within a small area of Kesennuma kanoue’s troops prayed for victory in their battles or and Minami-Sanriku. A street connected Koro- when Sakanoue’s enemy general was subjugated. mogawa and Kesennuma districts, which were both The fact that the enshrined deities for most of these headquarters of the Abe clan, and likely where shrines (three of five) were the Hachiman Deity troops from Tagajo passed through.

Accumulated Number of Shrines 120

110

100 Period V

90

80

70 1896 Meiji-Sanriku Tsunami Period IV 60

869 Jogan Tsunami 50 Period III

40 1611Keicho Tsunami 30 Period II

20 PeriodI

10

0 600 800 1000 1200 1400 1600 1800 2000 Year Fig.6 The time of founding and accumulated number of shrines.

235

(4) Period III (around 1090): founding of shrines Table 7 Founding of shrines and background. and background Period Shrine Year Background / Founder Sources B pp.390 Period III includes 14 shrines founded in the pe- MS02 801 War (Sakanoue) riod 1184–1206. The research results from docu- C pp.778 KJ01 807 War (Sakanoue) A pp.343 ments related to the history of the founding of A pp.318 MI17 807 War (Sakanoue) shrines are shown in Table 7. The results indicate I D pp.86 many cases where individuals based in the Tohoku KA53 807 War (Sakanoue) A pp.286 A pp.273 region founded shrines by transferring deities RT27 807 War (Sakanoue) (kanjo). E pp.167 IS75 809 Unknown B pp.50 In contrast to Periods I and II when most found- IS73 1043 Unknown B pp.47 ers belonged to the Imperial Court, many founders B pp.371 KE37 1051 War (M., Yoriyoshi) of shrines in this period were based in the Tohoku E pp.481 region. As indicated in Table 7, which provides a B pp.364 KE25 1064 War (M., Yoriyoshi) F pp.513 summary, 13 of 14 shrines were founded by identi- G pp.842 II fiable individuals, 10 of these by individuals based KE54 1064 War (M., Yoshiie) G pp.834 in Tohoku. Founders ranged from people in author- B pp.377 G pp.830 ity to individuals in lower classes, for example, MS20 1064 War (M., Yoshiie) H pp.400 from the wife of a medieval village lord to settlers I pp.78 from outside and local residents. The reasons for IS77 1064 Unknown B pp.48 founding shrines were also unique. In contrast to RT03 1184 Lord of a local castle A pp.273 Periods I and II, when shrines were founded in as- KA54 1187 Unknown J pp.514 B pp.346 sociation with wars, many were founded during this IS32 1188 Refugee from Fujiwara, Y. K pp.528-530 period for personal reasons (e.g., a shrine dedicated KU07 1189 Servant of M., Yoritomo A pp.343 to its founder’s favorite horse, a faithful individual MS19 1190 Fujiwara, Hidehira I pp.79 B pp.325 who transferred the deity, or an individual claiming MI28 1190 Wife of a local lord to have received a sign in a dream). There were also D pp.86 A pp.269 RT23 1190 Servant of M., Yoshitsune cases where samurais, who were granted land ten- E pp.171 ure around the period of war, transferred deities B pp.352 IS25 1190 Fujiwara, Hidehira K pp.524-525 from their homelands. There were as many as 13 III En-no-Raijo; Mountain- A pp.262 cases of “kanjo,” the ceremonial transfer of a di- OF49 1191 eering ascetic L pp.123 vided tutelary deity to a new location, of the 14 B pp.376 KE02 1192 Lord of a local castle shrines in this period. Judging from these factors, G pp.843 Immigrant from Heike this period marked the beginning of when shrines IS17 1198 M were founded “by individuals based in Tohoku for Clan A pp.267 personal reasons and by transferring a deity.” RT44 1200 Local resident E pp.147-149 Wife of a lord of local G pp.833 KE57 1201 (5) Period IV (around 1340): founding of shrines castle N pp.909-910 A pp.267 and background RT43 1206 Lord of a local castle Period IV includes five shrines founded in the pe- E pp.151-152 IS43 1336 Local lord B pp.39 riod 1336–1346. The research results from docu- TA01 1337 Daughter of a local lord A pp.267 ments related to the history of the founding of IV MS08 1337 Servant of Local Lord B pp.39 shrines are shown in Table 7. Since there is not IS47 1341 Unknown B pp.39 much information indicating the founders or back- NO04 1346 Unknown O pp.778 ground of the founding of these shrines, details are Sources; A28), B29), C30), D31), E32), F33), G34), H35), I36), J37), K38), unclear. Notwithstanding, the same trend from Pe- L39), M40), N41) , O42) riod III seems to prevail, based on the written rec- ords on Omiya and Mishima Shrines and founding retrieved, it is likely because the population in the by locally influential persons or their families coastal areas of the Sanriku region expanded and the lives of people in these villages stabilized conse- (6) Period V (around 1550–1750): founding of quent to Tokugawa’s governance. As such, there shrines and background were fewer conflicts during this period. Not many Period V is approximately 200 years, from around written records remain that indicate the founders of 1550 to 1750. Shrines were frequently founded shrines (18 of 43), which demonstrates the possibil- throughout this period—43 in total. Although the ity that an increasing number were founded by definite reasons behind the increase in the number non-notable persons during this period. Those with of shrines founded during this period could not be records of names indicate that the founders were

236 families or individuals likely from the local villages. culture of the time, were introduced into Tohoku through Hiraizumi. Suga Shrine, founded by Fuji- (7) Times when shrines were founded and back- wara no Hidehira with a transferred deity, is located ground near Shizugawa Bay and Shizugawa River (MS19). The preceding sections demonstrated that shrines This area was a shoen (manor) of the Fujiwara clan, have been founded for different reasons from Period where the fourth son of Fujiwara no Hidehira was III (around 1190) onwards. sent and where he lived in Asahitate Castle. “Gold- According to Okada43), the religious beliefs of en Street,” a street leading to Hiraizumi, where peo- average citizens drastically shifted from ancient to ple passed through to deliver their nengu (gold medieval mindsets in the mid- to late Heian period. and/or horses as land tax) was located here. As de- Whereas people once worshipped ujigami (clan dei- scribed earlier, kanjo-style shrines were first seen in ties) or tochigami (local deities), the new trend of the cities connected to Hiraizumi, the gateway to selecting and transferring deities from outside be- Tohoku at that time, indicating that the kanjo style, came widely accepted. This corresponds to the phe- the trend for founding shrines, reached the Tohoku nomenon identified in Period III. In the ancient be- region from Kyoto. lief system in various regions across Japan, people Thus, the religious beliefs in the coastal areas of worshipped the local deity, which in many cases the Sanriku region transformed from the ancient into was a mountain. Thus, the place where the deity the medieval style around the mid-12th century from should be, most often a unique natural environment, the areas connected with Hiraizumi as the starting was enshrined. Believers did not need to make rou- points. As the latter style of shrines were founded by tine pilgrimages to the shrine, and they visited only transferring deities from outside for prayer for per- on special occasions such as for religious ceremo- sonal reasons or worldly benefits, the shrines needed nies. However, from the medieval period onwards, to be located closer to people’s villages. Therefore, people opted to transfer their deities from the out- there was most likely a significant change in the side. Given their motivation for doing so (e.g., process of selecting shrine sites during this period. prayers for worldly benefit), it was considered im- This section demonstrates the local process of portant to make pilgrimages to the shrine, which led transformation of the religious beliefs by referring to the founding of shrines near the villages transfer- to historical records. The result corresponds to the ring deities. This explains that this transformation of general tendency in Japan which Okada pointed out people’s religious beliefs, which occurred around and the local details have become clear for the first the mid-12th century in the coastal areas of the time. The following section demonstrates an analy- Sanriku region, likely affected the selection of the sis on the relationship among the shrine’s founding shrine’s site. era, their location’s topographic features, and the In addition, of the 14 shrines founded in Period damage from tsunami. Also, there is no previous III, four were located within Hirota Bay and three at research to demonstrate this kind of comprehensive the mouth of the Oppa River. Only one was located analysis. elsewhere in the area. To elaborate, two of the four shrines located in Hirota Bay were confirmed to (8) The difference in topographical features and have been founded to enshrine transferred deities as damage between ancient and medieval belief guardians for its founders, who were granted land system tenures after General Yoritomo Minamoto’s war a) Topographical features of shrines covered by victory. Of the shrines founded around the mouth of the ancient belief system the Oppa River, one was founded by Fujiwara no This section presents the difference in ancient and Hidehira from Hiraizumi and another by a group of medieval belief systems in relation to topographical fleeing warriors from Hiraizumi, both with trans- features where shrines were located. The author ferred deities. In other words, about half the shrines found written records on when 138 of all 466 were founded by those familiar with the culture shrines were located at their current places. The 16 worshipped by central authorities such as the Mi- oldest shrines are located at their current places in namoto or Fujiwara clans. The coast of Hirota Bay the era of the ancient belief system (i.e., before and the mouth of the Oppa River were accessible to 1100). We were able to identify 13 shrines of the 16 and from Hiraizumi (a city that reached peak pros- through our field survey. perity in 1150 but was taken by Minamoto no Yori- As seen in Table 8, the topographic character of tomo in 1189). Thus, the culture of the central au- shrines covered by the ancient belief system is sim- thority was first introduced to the Tohoku region by ple as 12/13 is T, S, and I type, which is a moun- land (on Imaizumi Street) and water (Oppa River). tain-related geography. A folklore in Yokoya- These medieval beliefs, which characterized the ma-Hachimangu, which is I type, states that the hill

237 Table 8 Topographical classification of the shrines which in which the shrine is located, was formed by a located at the current positions in the ancient belief flooding of Hei river once upon a time. That indi- system era. cates the nuance of the ancient belief system to typ- Year Shrine Shrine FoundingYear Located at the Topographical ically see sanctity in natural environment. Current Posi- Classification The only exception is the V type of Nariyoke tions Shrine. This shrine was a temporary shrine of MI15 680 680 I Murone Shrine in Orikabe village, founded on the KE12 718 718 V IS67 748 748 S way to Orikabe during transferring the deity. That is OF23 783 783 S why the place is not on a mountain but deep within MS02 801 801 T a valley, because the place, by the old street, made it MI17 807 807 T much more convenient to found a temporal shrine KA53 807 807 T than it was to do it on a mountain. Thus, the shrines RT27 807 807 Unknown IS75 809 809 Unknown covered by the ancient belief system basically locate MS15 877 877 S where related to a mountain, such as ridges or single RT41 993 993 T hills. IS73 1043 1043 S b) Topological features of shrines located in the KE37 1051 1051 T medieval belief system KE25 1064 1064 T KE54 1064 1064 S The shrines included in the era of the medieval IS77 1064 1064 S belief system total 122. The author listed the oldest 16 of 122 in Table 9. The most typically different Table 9 Topographical classification of the shrines which point is the apearance of the F type, with as many as located at the current positions in the medieval be- three in only 10 years from 1190. That indicates lief system era (only 16 oldest shrines). that the change in the belief system in shrines Year Shrine Located at the Topographical caused drastic change in the selection of the sites. Shrine FoundingYear Current Posi- Classification Figure 7 presents the periods in which the shrines tions were located at their current places sorted by the MI18 1150 1150 T topographical classification. The F type appears as MI29 1157 1157 S RT03 1184 1184 T early as the start of the medieval era and has con- KA54 S th 1187 1187 stantly grown in number from 15 century onwards. KU07 1189 1189 F th C type comes later in the 14 century. That is, in the MS19 1189 1189 S era of the medieval belief system, shrine locations IS48 Unknown 1189 M became more diverse because people expected MI28 1190 1190 F OF49 1191 1191 S stronger relevance between the characteristics of the KE02 1192 1192 T enshrined deities and the location of the shrine, as IS17 1198 1198 T they began seeking a place to conduct prayers for RT44 1200 1200 S private or worldly benefits. KE57 1201 1201 F c) Difference in damage between ancient and IS80 Unknown 1201 S RT43 1206 1206 T medieval belief system YA05 1235 1235 V The “destroyed” shrines in the ancient belief sys- 100%0% tem era is only one out of 16 (6%) and that of the 1901- medieval belief system era is 12 out of 122 (10%). 90% 1801-1900 The shrines located at the current places in the ear- 80% lier era are slightly safer. 1701-1800 70% The only shrine that was “destroyed” was the Ki- 1601-1700 tano Shrine Imaizumi Tenmangu (RT41). There are 60% 1501-1600 two patterns of records about the founding year of 50% this shrine: one is 993 and the other one is 1478. 1401-1500 40% The records that adopt 993 also describe that the 1301-1400 shrine was constructed by a famous constructor, 30% 1201-1300 Dokan Ohta (1432-1486). The author adopts the 20% earlier one because 1478 seems to be a year when 1101-1200 10% the shrine was reconstructed at the same place -1100 where it had been. However, if the year 1478 was 0% the correct founding year, the difference between T S V F C M I the ancient and medieval belief systems would be Topographical Classification clearer. Fig.7 Periods within which shrines located at the current places sorted by the topographical classification.

238

6. LOCATIONS OF DESTROYED sition, meaning it has survived more tsunamis, the SHRINES COMPARED TO PREVIOUS safer its location. TSUNAMIS (2) Inundation from previous tsunamis (1) Time at the location of shrines and damage This section presents whether “damaged” shrines The author has a working hypothesis that the old- had been hit by previous huge tsunamis by examin- er a shrine has been located at its current position, ing the inundation area of those tsunamis. However, meaning it has survived more tsunamis, the safer its as the inundation area of the Keicho Tsunami of location. To test this hypothesis, we analyzed the 1611 has not been specified yet, shrines which were relationship between the length of time a shrine has founded or moved to a current location in periods B been located at its current position and the damage it and C were checked by only Meiji-Sanriku Tsunami has sustained. (1896) and the second largest, Syowa-Sanriku Tsu- The author found written records on when 138 of nami (1933). The inundation areas of past tsunamis all 443 shrines investigated in this research were were specified by the report of the Urban Planning located at the current positions. Figure 8 shows the relation between the length of time a shrine has been Table 10 Year “destroyed” shrines located at the current posi- located at its current position and the accumulated tion. Reloca- number of shrines. The timeline is classified into Found- Background for Period Shrine ingYear tion Relocation four periods (A–D) according to three previous Year large-scale tsunamis, namely the Jogan Tsunami RT41 993 - (869), Keicho Tsunami (1611), and Meiji-Sanriku KE02 1192 - B RT32 Tsunami (1896). We examined the historical back- 1328 - NO04 1346 - ground of “destroyed” shrines for each period (Ta- IS45 1490 - ble 10). KA14 1625 - There are nine shrines classified into the A peri- MI35 1659 - od, none of which were “destroyed” by the tsunami OT13 Unknown 1741 Because of fire (1724) C ON18 1787 - in 2011, 50 including 5 (10%) “destroyed” into the KE42 1801 - B period, 62 including 6 (10%) “destroyed” into the Moved from top of moun- YA04 C period, and 17 including 2 (9%) “destroyed” into Unknown 1854 tain to the current position the D period. Moved from top of moun- tain to the mountainside This result indicates as inappropriate the working (1854) IS49 hypothesis, according to the historical materials, that D 782 1919 Suffered from Meiji-Sanri- the older a shrine has been located at its current po- ku Tsunami, then moved to the current position YA06 1986 -

Accumulated Number of Shrines Period D 140140 Period C

130 LEGEND × : DESTROYED 120
: DAMAGED 110 ● : UNDAMAGED Period B 100

90 1611 Keicyo Tsunami 80

70

60 Period A

50 869 Jogan Tsunami 40 1896 Meiji-Sanriku Tsunami 19331933 SyowaSyowa-SanrikuSanriku TsunamTsunami 30

20

10

0 600 800 1000 1200 1400 1600 1800 2000 Year Fig.8 The length of time in which shrines have been located at the current position and accumulated number of shrines.

239

Period B

RT41 Kitano Shrine Imaizumi Tenmangu RT32 Yasaka Shrine KE02 Kamo Shrine

NO04 Watatsumi Shrine Period C

MI35 Ishihama Shrine

YA04 Osugi Shrine

KA14 Shirahama Shrine

ON18 Isuzu Shrine OT13 Shinmeidou Period D

NOTE - Aerial Photos : The author illustrates the location of shrines and traces the inundation lines on the photos in the report of the Urban Planning Section, Chief Cabinet of Minister of Home Affairs - Geographical Maps : The author illustrates the location of shrines and traces the inundation line on the disaster prevention maps of each prefectures or municipalities. IS49 Shinzan Shrine Fig.9 Relationship between the “destroyed” shrine’s position and inundation area of previous tsunamis.

240

Section, Chief Cabinet of Minister of Home Af- tended to be located near the low-lying coastline or fairs44) and, if not mentioned in the material, by the on flatlands. disaster prevention maps of each prefectures or mu- Second, the investigation on the size attributes of nicipalities (Fig.9). shrines indicated a minimal relationship with a) Period B: relationship between destroyed damage except for tiny shrines, which were not shrines and previous tsunamis buildings, just boxes, that tended to be located in The relationship of one shrine that stands in Og- flatlands. atsu-Ohama with the inundation area could not be However, the investigation on the enshrined deity identified because there were no materials found. In indicated a relationship between location and the remaining four shrines, only Watatsumi Shrine damage. In some deities, Susanoo, Inari, Kompira, (NO4) stands in the inundated area of the previous Munakata descents of shrines for example, the tsunami. That is the shrine that presumably experi- location tendency can be explained by those deities’ enced a tsunami in the past and was still located at characters. the same place. The reason seems that it needed to b) General characteristics of shrine locations in be located near the sea as its enshrined deity was a the coastal areas of the Sanriku region deity of the sea. The characteristics of shrine locations have b) Period C: relationship between destroyed greatly changed since the middle of the 12th century. shrines and previous tsunamis This research indicated that shrines were mostly The author was able to specify the inundation ar- positioned along the ridge line during the time of the eas of five shrines except for the one in Saichi, ancient belief system before the aforementioned Kesennuma. Among the five shrines, only Ohsugi turning point. Such locations were selected because Shrine (YA4) stands in the inundated area of both the connection between the shrine and the natural the past tsunamis. Ishihama Shrine (MI35) was in- environment was prioritized over connection to the undated at least once. Shirahama Shrine (KA14) village. This is likely because believers were not stands on the inundation line of both tsunamis. expected to embark on routine pilgrimages. In addi- c) Analysis tion, shrine locations became more diverse during The preceding sections demonstrate that shrines the medieval period from the mid-12th century on- which experienced past tsunamis and were “de- wards, when the culture of “kanjo,” the ceremonial stroyed” by the tsunami in 2011 were limited to transfer of a divided tutelary deity to a new location those that had relation between the enshrined deity according to deity’s character, was widely intro- and the location, such as the Watatsumi Shrine who duced to the area. Shrine locations became more is the deity of the sea. The only shrine whose rela- diverse because people expected a stronger rele- tionship between the deity and the location could vance between the characteristics of enshrined dei- not be identified by the author was the Ohsugi ties and the location of the shrine, as they began Shrine (YA04). seeking a place to conduct prayers for private or On the other hand, those shrines that did not ex- worldly benefits. This is a general site selection perience previous tsunamis, did not have any incen- characteristic for shrines. tive to relocate to other places, even if the shrine c) Local characteristics of shrine locations in the barely survived. That was supposed to be a reason coastal areas of the Sanriku region for them to be “destroyed” by the tsunami in 2011. This research demonstrated that the shrines com- That is, the experience with tsunamis affects the pletely destroyed or washed away during the tsuna- local site selection system for shrines. mi in 2011 were not located in the inundated areas of previous tsunamis, with a few exceptions. This explains that the locations of shrines in coastal areas 7. CONCLUSION of the Sanriku region align with the aforementioned process of shrine location selection, which is also (1) Conclusion dictated by the unique history of disasters that occur The findings of this paper can be summarized as in each area. This is a local site selection character- follows: istic for shrines. a) Damage to shrines in the tsunami in 2011 Of the shrines located in the villages of the (2) Discussion on regeneration Sanriku coastal area focused on in this research, 60 In my opinion, the argument of designating (14%) were completely destroyed or washed away shrines as evacuation shelters needs to be consid- during the tsunami in 2011. Further, shrines were ered carefully. Given the findings of this research, a generally located on ridges of mountains less likely number of shrines would not be safe if a tsunami to be affected by tsunamis and the destroyed shrines with higher reach than previous ones occurs. In

241 emergencies, while it is important to refer to the nohata village/ IW : Iwaizumi town/ MI : Miyako city/ YA : hazard map, it is more important to acknowledge Yamada town/ OT : Otsuchi town/ KA : Kamaishi city/ OF : that the cardinal rule is to “evacuate to a higher Ofunato city/ RT : Rikuzen-takata city / KE : Kesennuma city/ place as possible.” MS : Minami-sanriku town/ ON : Onagawa town/ IS : Ishi- On the other hand, to understand the historical nomaki city background of the shrine site will add to the persua- Name of Shrine : includes abbr. listed below. siveness and rationality of the idea of designating S: Shrine/ H: Hachimangu/ D: Daimyojin/ M: Myojin or My- the shrine as a key location in regeneration plan- ojinja/ T: Tenmangu/ DG: Daigongen/ A: Amaterasu-Mioya ning. Although it is difficult to designate shrines, Size Classification : includes abbr. listed below. L: Large/ M: Medium/ S: Small/ T: Tiny/ U: Unknown many of which are religious institutions, as part of Enshrined Deity : uses marks listed below. public regeneration planning, it is still possible to A: Susanoo descent/ B: Kumano descent/ C: Hachiman descent/ use their historical background in the planning pro- D: Inari descent/ E: Amaterasu descent/ F: Kompira descent/ G: cess. Munakata descent Finally, I want to suggest the possibility of utiliz- Damage Classification : uses marks listed below. ing the shrine’s site selection characteristics as a A: Destroyed/ B: Damaged/ C: Undamaged/ D: Unknown way to hand down the tsunami experience to the next generations. Shrines in individual villages can 1 2 3 4 5 6 7 8 9 10 11 KJ1 Wakamiya H. Natsuicho 26.5 1,510 T L X 807 1953 C only be seen as “spots.” If we look at all the shrines KJ2 Shioiri Inari D. Natsuicho 36.6 774 T M D - - C KJ3 Kompira S. Minatomachi 41 669 T M F 1718 1718 C located around a bay, a group of shrines can become KJ4 Fushimi Inari D. Niida 5 1,150 F S D - - C KJ5 Ebisu S. Niida 5.2 1,224 F S X - - C a line. If we select the correct ones, the line of KJ6 Kuji Inari S. Osanaicho 4.9 1,709 F L D 1658 1658 C KJ7 Suwa M. Osanaicho 4.9 1,643 F L X 1189 1189 C shrines would extend slightly above the tsunami KJ8 Tamanowaki S. Osanaicho 9.7 54 C S X - - C KJ9 Ebisu S. Kosode 6.5 39 C S X - - A inundation lines of the largest previous tsunamis. If KJ10 Junishin S. Kosode 10.2 83 S S X - - C KJ11 Yakata M. Kuki 31.3 48 C U X - - A some actions, for example, religious ceremonies KJ12 Ebisu S. Kuki 12 64 T T X - - A KJ13 Ebisu S. Kuki 41.3 250 F M X - - C ideally, are begun with these shrines in collabora- NO1 Inari D. Nodamura 30 1,204 S M D - - C NO2 Atago S. Nodamura 24.6 1,054 T L X - - B tion, this would be a greater help in transmitting NO3 H. S. Nodamura 18 997 F S C - - C NO4 Watatsumi S. Mikkaichiba 12.6 88 C M X 1346 1346 A tsunami experiences to descendants than tangibles NO5 Tamagawa S. Tamagawa 20 81 T M E - - B FD1 Goushi S. Ootanabe 25.9 141 T M F - - C such as monuments. The points described above TA1 Omiya S. Raga 39.3 145 S M E 1337 1337 C TA2 Kumano S. Hiraiga 48.7 248 T S B - - C explain the benefit of creating a network of shrines TA3 Sonsya Daijingu Simanokoshi 21.9 138 S M E 1240 1240 B TA4 Unknown Simanokoshi 33.2 139 M S X - - C as long-lasting common key places to share disaster IW1 Hakusan S. Omoto 29.1 1,047 T S X - - C IW2 H. S. Omoto 34.4 458 V L C - - C experiences. IW3 Kumano S. Omoto 30 623 T M B - - C MI1 Dewa S. Taro 16.5 173 T M D - - B MI2 Kumano S. Taro 32.2 642 T S X - - C MI3 Hie S. Taro 92 704 M U X - - C ACKNOWLEDGMENTS: This research was MI4 Unknown Matsuzuki 50.6 1,232 S S X - - C MI5 Kibune S> Onappe 19.7 604 S S X - - C supported by the Japan Society for the Promotion of MI6 Wakamiya H. S. Kuwagasaki 26.3 320 T S C - - C Science (JSPS) KAKENHI Grant Number MI7 Kumano S. Kumanocho 32.8 211 T L X 1539 1934 B MI8 Kompira S. Kamicho 34.5 197 S U X 1814 1814 C JP25870189. This research was the result of the MI9 Minato Osugi S. Kouganchi 50.4 226 T M X - 1933 C MI10 Atago S. Atago 58.3 711 T L X - 1944 C joint research with CSIS, the University of Tokyo MI11 Hankan Inari S. Sawada 31 1,088 S S X - - C MI12 Haguro S. Yokomachi 37 1,386 T M X - - C (No.394) and used the following data: Zmap-Town MI13 Ishizaki S, Yokomachi 40.9 1,387 T M X - - C MI14 Inari D. Kamosakicho 34 2,146 T S D - - C II 2008/09 Shape version (Iwate and Miyagi prefec- MI15 Yokoyama H. Miyamachi 21.8 1,898 I L C 680 680 C MI16 Kompira S. Miyamachi 21 1,865 I S F - - C tures) data provided by Zenrin Co., Ltd. MI17 Koyamada S. Koyamada 25.6 2,502 T M X 807 807 C MI18 Fujiwarahiko S. Fujiwara 26.1 537 T M X - 1961 C MI19 Yagizawa S. Yagizawa 17.3 1,584 T S X - - C MI20 Kumano-San S. Sokeioki 15.1 127 I M B - - C MI21 Ukishima S. Sokei 2.3 125 F T X - - A MI22 Inari D. Kambayashi 16.7 51 I L D - - C APPENDIX : DATA OF THE SHRINES MI23 Inari S. Takahama 14.4 272 T M D 1730 1730 C MI24 Kanahama Inari S. Kanahama 22.6 240 T L D 1661 1661 B MI25 Kumano S. Tsugaruishi 28.1 192 V S B - - B MI26 Sei-ishii Inari S. Tsugaruishi 16.7 670 T M D - - C The list covers the research data of all 466 MI27 Hankan S. Tsugaruishi 40.3 1,557 T U X - - C MI28 Komagata S. Tsugaruishi 20.1 1,989 F M X 1190 1190 C shrines. MI29 Akamae H. Akamae 50.6 1,012 M L C - - C MI30 Hakusan S. Shirahama 20.3 483 F S X - - C MI31 Unknown Uiso 19.2 150 S U X - - A MI32 Daijingu Otobe 46.1 591 T L E - - C [LEGEND] MI33 Shinshi S. Nishi-Odate 24.4 504 T S D - - B MI34 Chikei S. Chikei 41.4 303 T L X - - C Heading of Table : MI35 Ishihama S. Ishihama 16.9 73 S L X 1659 1659 A MI36 Kagura S. Kawashiro 55.8 210 S M X - - C 1. ID Number/ 2. Name of Shrine/ 3. Name of District/ 4. Alti- YA1 Unknown Osawa 10.7 249 S M X - - C tude/ 5. Horizontal Distance from the Nearest Coastline/ 6. YA2 H. Osawa 14.9 262 T M C - 1761 C YA3 Nagahama S. Osawa 8.8 121 T L X 997 1966 B Topographic Classification/ 7. Size Classification/ 8. Enshrined YA4 Osugi S. Kitahamacho 2.9 164 F L X - 1848 A YA5 H. S. Yamada H. Hachimancho 28.9 500 V L C 1235 1235 C Deity/ 9. Founding Year/ 10. Year to Locate in a Current Place/ YA6 Hotate Kaido S. Kawamukai 3.9 351 F T X 1986 1986 A YA7 Kompira S. Sakaidacho 19.9 399 T M F - - B 11. Damage Classification YA8 Suitengu Sakaidacho 9.7 224 S M X - - C YA9 Unknown Oura 2.8 49 C T X - - C ID Number : consists of municipality name abbr. listed below YA10 Karogatake S. Oura 15 237 F M X - - C YA11 Koyadori S. Koyadori 33.8 154 S M X - - C and number. YA12 H. Funakoshi 22.9 195 T M C - 1717 B KU: Kuji city/ NO : Noda town/ FU : Fudai village/ TA : Ta- YA13 Tateyama H. Funakoshi 13.3 96 I M C - - B YA14 Suwa S. Funakoshi 14 157 T L X 1655 1655 C

242

YA15 Shiogama S. Funakoshi 41.5 132 T M X - - C OF38 Sanko nari S. Tate Shita 44.9 2,378 S S D - - C YA16 Kumano S. Funakoshi 19.1 149 T S B - - C OF39 Kibune S. Tsunozawa 19.8 1,893 T M X - 1973 C OT1 A. S. Kirikiri 21.2 249 T L E - - B OF40 Nokyo Inari S. Tate Shita 14.3 1,492 T S D - - B OT2 Inari D. Kirikiri 14.6 396 F M D - - B OF41 Shinzan S. Tomizawa 30.3 1,267 T M X - - C OT3 Kompira S. Kirikiri 18.4 635 F M F - - B OF42 Yasaka S. Sasazaki 38.9 493 T L A - - C OT4 H. Akahama 31.3 345 S L C 1733 1733 B OF43 Kamo S. Sasazaki 41.9 496 T L X - 1955 C OT5 Kumiai S. Ando 7.6 165 F T X 1940 1940 A OF44 Kumano S. Sasazaki 38.1 468 S S B - - C OT6 Otsuchi Inari S. Ando 27.5 447 T L D - 1720 B OF45 Gassan S. Shimofunato 56.8 415 T L X - - C OT7 Komagata S. Ogaguchi 9.9 2,450 S M X -- A OF46 Matsukura S. Funagawara 8.9 189 S M X - - C OT8 Hachiman S. Otsuchi 37.8 2,507 S S C 1644 1644 C OF47 Kompira S. Kamisaka 21.2 56 S M F - - C OT9 Kumano-San S. Ogaguchi 38.1 2,146 T S B - - C OF48 Yasaka S. Kamisaka 23.4 117 S S A - - C OT10 Fushimi Inari D. Otsuchi 9.7 973 T S D - - A OF49 Kumano S. Kamisaka 25.3 101 S L X 1191 1191 C OT11 Kozuchi S. Sukacho 6.6 245 F M X - - A OF50 Kumano S. Nakamori 13.4 205 S L B 1256 1256 C OT12 Kozuchi S. Kamicho 12.4 990 V L X 829 1629 B OF51 Hachidai Ryujin Konakai 10.5 69 S L X - - C OT13 Sinmeidou kamicho 9.8 1,095 F M X - 1741 A OF52 Hachiman S. Takashimizu 18.5 138 S M C - - B OT14 Kozuchi D. Usuzawa 43.2 3,027 T S X - - C RT1 Kumano S. Yachidate 41.4 251 S U B - - C KA1 Kanzeon S. Murohama 13.6 344 T L X - - B RT2 H. Yachidate 33.3 79 M L C 1385 1385 C KA2 Murohama Inari S. Murohama 10.2 161 T T D - - A RT3 Jagasaki S. Yachidate 24 36 T T X 1184 1184 C KA3 Inari D. Katagishi 6.3 504 F T D -- A RT4 Kumano S. Tadade 10.6 83 F M B - - A KA4 Tate Inari S. Katagishi 29.8 601 T M D 16911691 B RT5 Inari S. Tadade 7.8 127 F S D - - C KA5 Inari S. Katagishi 20.4 730 0 U D - - U RT6 Kompira S. Tadade 14.6 62 I U F - - A KA6 Rokusan S. Unosumai 14.3 1,664 T M X 1710 1710 B RT7 Kompira S. Nagahora 49.9 383 M S F - - C KA7 Hakusan S. Unosumai 31.2 1,446 0 U X 1715 1715 U RT8 Niwatari S. Nagahora 42.3 208 S U X - - C KA8 Unknown Unosumai 28.6 1,531 V S X - - B RT9 Yasaka S. Maehanagai 31.4 94 I L A - - B KA9 Unosumi S. Unosumai 33.1 1,558 V L D 1690 1690 B RT10 Kompira S. Maehanagai 22.9 209 T L F - - C KA10 Sirahada Gongen Unosumai 13.9 1,107 S M X - - A RT11 Unknown Iwakura 34.8 325 S S X - - C KA11 Unknown Nebama 18.2 377 0 U X -- U RT12 Bato Kanzeon Iwakura 19.9 177 T M X - - C KA12 Fuouki S. Nebama 35.5 205 T M X -- B RT13 Hudoson Nezaki 23.9 166 F S X - - C KA13 Kamimae Inari Hakozaki 17.5 589 V S D - - C RT14 Tsuruju S. Nezaki 14.9 117 S L G 1557 1557 C KA14 Shirahama S. Hakozaki 7 271 F T X 1625 1625 A RT15 Kamo M. Atsumari 31.5 49 T M X 1704 1704 C KA15 Hakozaki S. Hakozaki 8.2 28 C M C 1677 1677 B RT16 Ugakonshin Tomari 14.9 285 F S X - - B KA16 Hoshinomiya S. Shirahama 22.4 180 T S X - - B RT17 Kompira DG. Tomari 15.5 93 T S F - - C KA17 Sanganjima S. Karijuku 36 70 T M G 17151715 B RT18 Atago S. Tomari 20 34 I M X - - C KA18 Unknown Kuwanohama 15.6 49 T M X - - A RT19 Osugi S. Tomari 16.4 27 I M X - - C KA19 Sujikawa Inari S. Kuwanohama 26.3 190 V S D 1825 1825 C RT20 Daisuijin Koshida 30.1 78 T S X - - C KA20 Tenno S. Ryoishi 71.4 111 0 U A -- U RT21 Komagata S. Usozawa 28.5 129 T S X - - C KA21 Itsukushima S. Ryoishi 40.5 248 T M G 17481748 B RT22 Yakumo S. Yanoura 17.9 169 F S X - - C KA22 Tukiyomi S. Mizuumi 24 179 T S X - - B RT23 Kumano S. Yanoura 17.7 50 T L X 1190 1881 C KA23 Takisai S. Shinhamacho 0 0 0 U X - - U RT24 Tahigashi S. Sawabe 117.2 686 0 U X - - U KA24 Imataki S. Higashimaecho 32.2 361 V S X 1704 1704 C RT25 Yamagami S. Moribana 38.5 1,157 M S X - - B KA25 Kompira S. Hmacho 28.9 371 S S F - - C RT26 Miyazaki Inari S. Miyazaki 21.2 1,277 T L D 1799 1799 C KA26 Inari S. Hmacho 0 0 0 U D - - U RT27 Hakusan S. Uenobo 0 0 0 U X 807 807 U KA27 Ozaki S. Hmacho 74.1 725 V L X 1699 1952 C RT28 Akiba S. Nakanishi 42.3 1,016 S S X - - C KA28 Yasaka S. Hmacho 0 0 0 U A - - U RT29 Akiba S. Naraimichi 40.1 454 S M X 1592 1824 C KA29 Kitsunezaki InariSya Hmacho 32.4 211 T S D - - C RT30 Yasaka S. Nakajima 27.7 372 T S A - - C KA30 Arayashiki Inari S. Tadagoecho 33.1 352 S U D - - C RT31 Yakuouzan Naraimichi 26.5 454 S M X - 1508 C KA31 Atago S. Tadagoecho 14.1 316 0 U X - - U RT32 Yasaka S. Gesyuku 13.2 590 T S A 1328 1328 A KA32 Atago S. Tadagoecho 28.7 276 T T X - - C RT33 Yama S. Yamanae 22.2 1,428 S T X - - C KA33 Yamazumi S. Otadagoecho 9.9 704 F T X - - C RT34 Mtomiya S. Nagasuna 28.4 1,071 S S X - - B KA34 Dewa-Sanzan Gobunsya Tadagoecho 6 384 F T X - - A RT35 Akiba S. Samukaze 21.3 1,353 F U X - - A KA35 Kumano S. Owataricho 4 684 F T B - - A RT36 A. S. Honmaru 38.8 1,617 T L E - 1941 B KA36 Isozaki Inari S. Omachi 26.5 780 T M D - - C RT37 Ichi Jingu Baba 4.8 1,500 F T X - - A KA37 H. Owataricho 42.3 1,152 T M C - - C RT38 Atago S. Nariishi 32.1 1,764 S S X - - C KA38 Matsubara S. Matsubaracho 41.3 554 T L A 1787 1787 C RT39 A. S. Sinmyoumae 15.9 4,369 S L E 1317 1317 C KA39 Fujita Inari D. Ureisicho 22.5 532 S S D - - C RT40 Komagata S. Katachi 24.1 4,574 S S X - - C KA40 Hakusan S. Ureisicho 26.9 530 T S X - - C RT41 Kitano S. Imaizumi T. Nakai 5.3 1,536 T L X 993 1478 A KA41 Bisyamon S. Ureisicho 20.2 415 S M X - - C RT42 Suwa S. Kesencho 21.3 1,275 T L X 1160 1758 B KA42 Inari Taisya Heita 17.5 233 T T D -- C RT43 Kajima S. Futsukaichi 32.8 164 T L X 1206 1206 C KA43 Ara S. Heita 19.1 348 S S X - - C RT44 Gassan S. Tsukiyama 41.3 605 S L X 1200 1200 C KA44 Ara S. Heita 7.3 179 F T D - - A RT45 Unknown Sugoroku 12.1 31 T S X - - C KA45 Suijin D. Heita 16 239 S T X - - C RT46 Yasaka S. Yougai 21.9 57 T M A - - C KA46 Tateyama S. Heita 36.9 147 S L X - - C RT47 Unknown Fuppushi 20.4 48 T M X - - C KA47 Ozaki S. Ozaki-Shirahama 30.8 99 T M X - - B KE1 Shiwa S. Takenosode 13.9 32 T S X - - C KA48 Sasu S. Sasu 12.8 50 T M X - - B KE2 Kamo S. Takenosode 6.4 112 T M X 1192 1192 A KA49 Roukai S. Kerobe 4.5 24 C M X - - A KE3 Unknown Tate 13.4 36 0 U X - - U KA50 Osugi S. Hongo 27.1 132 T L X -- C KE4 Takekoma S. Oohata 19.4 71 T S X - - C KA51 Osugi S. Hongo 38.9 196 T M X -- C KE5 Unknown Tadakoshi 63.3 349 0 U X - - U KA52 Yasaka S. Kojirahama 32.3 187 S M A - - C KE6 Yakumo S. Tadakoshi 34 308 T S X - - B KA53 A. S. Katagishi 22.5 318 T L E 807 807 C KE7 Uga S. Tadakoshi 28.1 372 S M X - - C KA54 Kumano S. Arakawa 24.2 427 S L B 1187 1187 C KE8 Unknown Ishihama 8.8 311 S M X - - C KA55 Bisyamon S. Oishi 5.8 53 C S X - - A KE9 Shinzan S. Akedo 33.3 111 M T X - - C KA56 Hachiman S. Oishi 50.9 210 T U C - - C KE10 Takino S. Kakehama 29.4 79 T S X - - C OF1 Owatatsumi S. Senzai 30.9 55 I M X - - C KE11 Kompira S. Kosaba 24.4 69 T M F - - B OF2 Benten S. Juniyaku 28.6 60 S L X 16171617 B KE12 Nariyoke S. Shibitachi 35.1 346 V M X 718 718 C OF3 Ichikishimahime S. Juniyaku 19.4 36 0 U G - - U KE13 Hayama S. Syukuura 30.8 92 S L D 1217 1626 C OF4 Shinzan S. Yoshihama 29.1 466 S L D - 1535 C KE14 Oritsuhime S. Nishimoune 9.4 236 S U X - - A OF5 A. S. Yoshihama 35.6 893 S U E - - C KE15 Mitake S. San-no-hama 17.5 74 T M X 1605 1605 C OF6 Hachiman S. Sugishita 30.3 366 T S C - - C KE16 Kompira S. Ni-no-hama 9.5 82 T S F - - B OF7 Shinzan S. Maeda 37.6 473 T L D 1298 1672 C KE17 Kumano S. Ni-no-hama 20.4 69 I S B 1763 1763 C OF8 Kompira S. Onizawa 12.6 33 C S F - - C KE18 Yama-no-Kami Kogoshio 23.6 63 T S X - - C OF9 Ichikishima S. Onizawa 33.1 220 S L G - - C KE19 Yama S. Kogoshio 21.6 149 S S X - - B OF10 Kumano S. Koishihama 51.6 354 S M B - - C KE20 Kumano S. Oura 30.7 154 S S B - - B OF11 Sakiyama D. Sunagohama 8.2 16 I M X - - C KE21 Itsukushima S. Oura 0.4 20 C M G 1746 1746 B OF12 Tateishi S. Kuwabata 23.2 50 T T X - - C KE22 Suga S. Namisaka 14.7 67 T M X - - B OF13 Unknown Nonomae 56.8 204 S T X - - C KE23 Yama-no-Kami Namisaka 4.8 315 F T X - - A OF14 Raijin S. Ryori 10.9 766 T L X -- C KE24 Iizuna S. Namisaka 8.5 458 S S X 1572 1572 B OF15 Akiba S. Ryori 23.2 274 T L X 1560 1560 B KE25 Hachiman S. Higashi Hachiman 7.2 1,560 T L C 1064 1064 C OF16 A. S. Ryori 16.4 81 S M E 1446 1714 B Mae OF17 Ichikishima S. Ishihama 6.9 45 C L G 1491 1787 B KE26 Kumano S. Higashi Nakasai 26.8 2,489 S S B - - C OF18 Unknown Attari 24 360 T M X - - C KE27 Tate H. S. Nishi Nakasai 55.6 2,271 M S C - - C OF19 Kibune S. Hokaguchi 27.3 121 T M X - - C KE28 Itsukushima S. Kurasoko 8 591 T M G - - C OF20 Atago S. Hokaguchi 40.9 66 M M X - - C KE29 Isuzu S. Sakanamachi 8.4 26 I L E 1427 1681 B OF21 Wakame S. Nagasaki 23.2 109 T S X - - C KE30 Igari S. Sakanamachi 11.2 25 I M X 1877 1877 B OF22 Unknown Nagasaki 26.2 102 S M X - - C KE31 Hiyoke Inari D. Youkamachi 33.5 381 S M D - 1635 C OF23 Ozaki S. Karasuzawa 12.5 73 S L D 783 783 B KE32 Murasaki S. Hamamiyama 12.4 306 S L X 1582 1648 C OF24 Unknown Takonoura 49.2 58 I U X - - C KE33 Kashiwazaki Sei Kannon Kashiwazaki 10 119 S M X 1582 1648 C OF25 Yaotome S. Shimizu 39.4 80 I U X - - C KE34 Ikkeijima S. Bentencho 1.3 190 F L G - - A OF26 Unknown Nagahama 30 230 0 U X - - U KE35 Kasuga D. Honmachi 17.9 852 T S X - - C OF27 Fushimi Inari Nagahama 6.8 245 F T D - - A KE36 Kumano S. Urata 31.2 943 F T B - - C OF28 Itsukusima S. Nagahama 60.6 111 I L G 17261902 C KE37 Hachiman S. Nakase 22.4 452 T L C 1051 1051 B OF29 Yasaka S. Oikata 9.4 177 T L A - - C KE38 Ozaki S. Ozaki 5 53 I L X 1648 1648 B OF30 Kumano S. Sawada 43.2 2,023 V S B - - C KE39 Hurumine S. Senganda 16.6 249 F M X - - C OF31 Yoshinomori S. Gongendo 19.6 3,166 T M X - - C KE40 Oyama S. Dainosawa 14.5 98 T S X 1616 1616 B OF32 Hayachine S. Kakinokizawa 17.3 2,910 T S X - - C KE41 Tsushima S. Tsukihoshiya 16.6 1,005 M T X - - C OF33 Goyo S. Morimotocho 17.8 2,795 S T X - - C KE42 Hachidai Ryujingu Kawara 0.6 33 C S X 1801 1801 A OF34 A. S. Kakinokizawa 55.6 2,783 M L E - 1709 C KE43 Unknown Arasawa 30.8 954 M T X - - C OF35 Tenjin T. Kakinokizawa 57.3 2,767 M M X - - C KE44 Akiba S. Okubo 42.1 946 T L X 1741 1741 C OF36 Komagata S. Kakinokizawa 18.2 2,701 S S X - - C KE45 Kumano S. Nagaisohama 15.7 108 S S B 1690 1690 C OF37 Nejiro H. S. Tate Shita 26.4 2,343 T M C 1555 1974 C KE46 Isozaki S. Nagaisohama 13 60 S M X 1626 1626 C

243

KE47 Kompira S. Hajikami 10.9 96 I L F 16351635 C ON4 Niwatari S. Takeura 0 331 0 U E - - U KE48 Bangyu Tenno Sya Hajikami 14.3 594 F S A - - A ON5 Isuzu S. Takeura 14.9 78 T L X - - C KE49 Kumano S. Horasawa 21.7 983 F M B - - C ON6 Yama S. Kirigasaki 9.9 132 F S X - - B KE50 Taduka S. Nonoshita 20.7 172 I M X - - C ON7 Isuzu S. Kirigasaki 21.7 89 S M E - - C KE51 Takigami S. Takine 21.3 492 T L X 1488 1488 C ON8 Tenman-Daijizai T. Ishihama 13.4 85 T M X - - B KE52 Hachiman S. Nichikado 21.8 64 T M C - - B ON9 Kumano S. Ishihama 28.3 205 T M A 1850 1850 B KE53 Hakusan S. Nichikado 22.7 319 S M X - - C ON10 Yamazumi S. Miyagasaki 10.5 126 T L X - - B KE54 Suga S. Sanya 19.9 324 S L A 1064 1064 C ON11 Unknown Miyagasaki 19.5 493 S T X - - C KE55 Kusushi S. Nakajima 13.4 405 I M X - - C ON12 Hakusan S. Onagawahama 16 375 T L X - - A KE56 Kusushi S. Nazawa 16.7 869 S S X -- C ON13 Kumano S. Washigamihama 55.4 393 T L A - - C KE57 Mitake S. Tsuyatateoka 36.5 1,407 F L X 12011201 C ON14 Niwatari S. Konorihama 12.9 63 F L X - - A KE58 Tate Okayama S. Tsuyatateoka 49.4 1,488 T S X - - C ON15 Niwatari S. Takashirahama 9.7 72 F M X - - A KE59 Yakumo S. Warabino 11.9 1,550 T M X - - C ON16 Niwatari S. Yokoura 16.5 60 T M X - - B KE60 Hachiman S. Sodeo 32 1,099 T L C 1584 1584 C ON17 Niwatari S. Oishiharahama 10.7 44 T M X - - B KE61 Akiba S. Kesaiso 6.8 125 F T X - - A ON18 Isuzu S. Nonohama 9.8 131 T M E 1787 1787 A KE62 Gion S. Utou 25.7 208 T L A - - B ON19 Shinmeigu Iigohama 13.6 58 S U X - - A MS1 Kusushi S. Utatsu 20 509 S S X - - C ON20 Kusushi S. Iigohama 16.4 64 S L X - - A MS2 Asahigaoka H. S. Tanoura 31 57 T M C 801 801 B ON21 Shinmeigu Tsukahama 28.2 136 T L X - - B MS3 Yakumo S. Tanoura 17.3 57 S L A - - B ON22 Shirahata S. Tsukahama 9 96 C M X - - C MS4 Siratori Inari D. Minaminosawa 37.9 444 M S D - - C ON23 Isuzu S. Tsukahama 11.2 79 S M E - - C MS5 Iizuna S. Ishihama 30.5 59 T S X - - B IS57 Myou S. Yoriisohama 41.5 41 T M X - - A MS6 Isuzu S. Minato 22 80 I L X - - B IS58 Yamagami Yoriisohama 34.7 218 T S X - - C MS7 Hachiman S. Inegafuchi 1.6 67 C T C - - A IS59 Kumano S. Yoriisohama 45.6 158 T L A - - C MS8 Mishima S. Isatomae 30 93 I M X 1338 1338 B IS60 Ama-no-Toyohiko S. Maeamihama 37.1 156 S M X - - B MS9 Kumaou S. Isatomae 30.5 426 M S X - - C IS61 Isuzu S. Sameura 13.9 81 T M E - - B MS10 Sarutahiko Ichi S. Isatomae 13.3 267 T M X - - B IS62 Kumano S. Sameura 4.2 45 I U B - - C MS11 Gongen Inari Niranohama 20.9 266 T M D - - C IS63 Niwatari S. Ooyagawahama 29.2 35 S M X - - B MS12 Hachiman S. Nishida 6.9 74 S M C -- C IS64 Hachiman S. Yagawahama 25.5 123 T L C 1569 1569 B MS13 Kusushi S. Matsuida 13.8 748 T M X 1569 1569 B IS65 Kompira DG. Tomarihama 27.6 73 S S F - - B MS14 Akiba S. Kitanomata 40.3 163 T S X - - B IS66 Isuzu S. Tomarihama 8.3 25 T M E 1486 1486 C MS15 Ara S. Sodehama 16.9 130 S L X 877 877 B IS67 Yanaki S. Shinzanhama 19.1 75 S S X 748 748 C MS16 Hakusan D. Omoricho 4.3 208 F T X - - A IS68 Kumano S. Ayukawahama 34.5 378 S L A - 1914 C MS17 Okuman Inari S. Omori 17.3 380 T M D - - B IS69 Inari D. Kugunari 16.6 346 V T D - - C MS18 Nishimiya S. Tennousan 34.5 486 T S X 1150 1150 C IS70 Hakusan S. Kugunari 7.6 86 T M X - - B MS19 Suga S. Tennousan 21.9 934 S S A 1189 1189 B IS71 Shinmeigu Kugunari 25.6 104 S M X - - B MS20 Kaminoyama H. Kami-no-Yama 19.4 647 S L C 1064 1971 C IS72 Kompira DG. Kugunari 21.1 78 S S F - - C MS21 Unknown Kami-no-Yama 17.6 483 T L X - - C IS73 Isuzu S. Obuchihama 20.3 96 S L E 1043 1043 C MS22 Furumine S. Kami-no-Yama 17.6 661 S L X 1895 1895 C IS74 Chokai S. Kyubunhama 0 60 0 U X - - U MS23 Inari S. Sukezukuri 9.9 762 T U D - - A IS75 Haguro S. Kyubunhama 0 417 0 U D 809 809 U MS24 Komori S. Komori 9.7 1,728 F T X - - A IS76 Kumano S. Oharahama 22.6 139 T L X 1591 1763 C MS25 Yama S. Okubo 17.5 517 V T X - - C IS77 Shinmeisya Koamikurahama 26 101 S M E 1064 1064 B MS26 Kurosaki Fudoson Kurosaki 33.5 185 S T X - - C IS78 Niwatari S. Fukkiura 23.5 119 S M X 1592 1592 B MS27 Isuzu S. Hoshiya 18.1 304 I M E -- B IS79 Isuzu S. Kitsunezakihama 11.9 79 S M E - - C MS28 Kaminari S. Hinata 20.7 1,799 S S X - - B IS80 Sei-ichii Inari S. Kitsunezakihama 29.4 69 S L D - 1201 C MS29 Tokura S. Hadenya 5.4 269 S L X 11571157 B IS81 Niwatari S. Takehama 28.2 84 S M X - - B MS30 Unknown Tsunomiya 6.5 132 S U X - - A IS82 Isuzu S. Makihama 12.1 126 T M E - - B MS31 Kusushi S. Kondo 36.4 100 S M X - - C IS83 Isuzu S. Kozumihama 31.8 110 T L E - - B MS32 Inari S. Fujihama 28.5 79 S S D - - C IS84 Hayamahime S. Ogihama 42.1 264 T L X - - C MS33 Niwatari S. Nagashimizu 35 144 S M X - - C IS85 Inari S. Samuraihama 0 81 0 U D - - U MS34 Unknown Nagashimizu 7.6 65 S S X - - A IS86 Isuzu S. Tsukiura 33.6 167 T L E - - C IS1 Isuzu S. Ooyubi 28.9 188 T S E -- C IS87 Unknown Momoura 8.8 209 S T X - - A IS2 Inari S. Koyubi 0 0 0 U D -- U IS88 Isuzu S. Momoura 24.9 149 S L E 1573 1573 C IS3 Kumano S. Aikawa 28.6 160 T M B - - C IS89 Hama S. Momoura 11.3 8 C U X - - A IS4 Inari S. Kodomari 0 213 0 U D - - U IS90 Izu S. Orihama 53 90 S M X - - C IS5 Yaama S. Omura 34.2 119 T L X - - C IS91 Hie S. Kotakehama 17.3 103 S M X - - B IS6 Isuzu S. Komuro 9.5 133 T M E - - C IS92 Iraka S. Watanoha 8.9 32 S S X - - C IS7 Kajima S. Shirahama 8.6 205 F M X - - A IS93 Isuzu S. Watanoha 13.8 103 T L E - - C IS8 Kasuga S. Shiotani 8.3 120 F M X - - A ON24 Isuzu S. Urajuku 14.3 149 T M E - - C IS9 Inari S. Tachigami 24.7 276 S M D - - B ON25 Yakumo S. Urajuku 35.2 154 M L A - - C IS10 Isuzu S. Tsukihama 44.2 1,159 T L E - - B IS94 Bisyamonten Shinohata 15 299 S M X - - C IS11 Kumano S. Yoshihama 31.1 1,913 T L B - - B IS95 Naganaki S. Oritate 9.9 79 S M X - - C IS12 Furumine S. Oppa 40.2 2,092 T M X -- B IS96 Hie S. Oritate 9 105 T M X - - C IS13 Tsuriishi S. Oppa 38.2 2,119 T L X -- B IS97 Kajima S Nagaru 10.9 116 S L X - - C IS14 Makioka S. Hashiura 30.2 4,247 T M X - - C IS98 Itsukushima S. Nagaru 7.9 31 C S G - - B IS15 Hojiki S. Hashiura 6.2 4,038 F M X - - B IS16 Atago S. Hashiura 7.4 4,568 F L X 1628 - C IS17 Kajima S. Hashiura 14.8 6,025 T M X 1198 1198 C IS18 Yakumo S. Tate Maehata 11 9,783 I L A 1238 1238 C IS19 Kumano S. Nakanodai Shita 6.2 10,162 F M X 1494 1494 C REFERENCES IS20 Sakura S. Makinosuyyama 19.9 8,218 T L X - 1711 C 1) Urban Planning Section, Chief Cabinet of Minister of IS21 Wakagi S. Takenosako 27.8 9,236 T S X - - C IS22 Kamo-no-Otono S. Kamozaki 38.6 6,922 T L X 899 - C Home Affairs : Sanriku Tsunami ni yoru Higai Chouson no IS23 Hachiman S. Fukuchi 9.4 5,916 S L C - 1908 C IS24 Yakumo S. Kokudo 9.7 4,338 T M A 1615 1615 C Fukkou Keikaku Houkokusyo (Report of Reconstruction IS25 Haguro S. Karasuyamori 74.8 3,946 S L D 11901819 C Plan on Towns and Villages from the Sanriku Tsunami), IS26 Haguroyama S. Harioka 9.6 3,367 0 U X - - U IS27 Hie S. Kamatani 10.4 2,035 T M X - - C Urban Planning Section, Chief Cabinet of Minister of IS28 Inari S. Kamatani 20.7 2,209 V S D - 1657 C IS29 Higo S. Kamatani 20.5 1,938 T S X - - C Home Affairs, pp. 42-43, 1934. IS30 Unknown Kamatani 4.9 1,972 F U X - - A 2) Yamaguchi, Y. : Nihon no Koyuseikatsu wo Motomete (In IS31 Yakumo S. Nagatsura 19.3 50 T S X - - B IS32 Kitano Tenman S. Nagatsura 33.1 90 M L X 11881689 C search of unique life in Japan), Sekai Bunko, Yamaguchi IS33 Unknown Ozaki 11.4 50 T S X - - C Yaichiro Sensyu, Vol. 6, 1972. IS34 Kusushi S. Ozaki 9.5 144 V L X 1683 1683 C IS35 Sanshi S. Naburi 42.4 203 S L X 14241872 C 3) Hitaka, K., Arima, T., Ikaruga, S., Sakai, T. and Hagishi- IS36 Yakumo S. Funakoshi 23.2 133 T L X - - B IS37 Kumano S. Funakoshi 34.1 113 S M X - - C ma, S. : A study on location characteristics of shrines IS38 Fushimi Inari D. Osu 23.4 50 T S D - - C drawn in the Local Meisho-Zue : An analysis of "Pictorial IS39 Takekoma Inari S. Osu 41.9 188 0 U D - - U IS40 Hachiman S. Osu 28.1 221 S L C 1358 1358 C Record of the Sights in Japan Fukuoka Prefecture", J. IS41 Isuzu S. Kumazawa 24.8 73 S L E - - C Arch. Plan. JIA, Vol. 597, pp. 93-100, 2005. IS42 Kajima S. Tatehama 14.6 149 S M X - - C IS43 Chinju Kitano S. Tatehama 39.5 122 S L X 1336 1336 C 4) Goto, R. and Nakaoka, Y. : A study on the settlements IS44 Unknown Ohama 15.3 60 0 U X - - U IS45 Hayama S. Kobama 12.2 60 C L X 1490 1490 A located in Saga lowland plains from the view point of set- IS46 Kumano S. Kojima 14.6 73 S M B -- B tlement-shrines, J. Arch. Plan. JIA, Vol. 551, pp. 197-203, IS47 Shiogama S. Myoujin 22 64 T L X 13411341 B IS48 Kumano S. Ogatsu 70.5 259 M S X -1189 C 2002. IS49 Shinzan S. Ogatsu 9 289 F L X 781 1919 A 5) Urasaki, S. : A study about relationship between asymmet- IS50 Kurusu S. Ogatsu 19.1 999 F T X -- C IS51 Isuzu S. Ogatsu 27.5 155 V M E -- A rical space composition and geomorphic characteristics of IS52 Isuzu S. Karakuwa 30.3 190 S M E - - C IS53 Yama S. Mizuhama 6.9 59 I T X - - C Kyu-Kensha shrines in Nara prefecture, J. JILA, No. 71, IS54 Sagyo S. Mizuhama 8.5 56 I L X - - C Vol. 5, pp. 779-784, 2008. IS55 Isuzu S. Wakehama 19.6 67 S L E - - C IS56 Yakumo S. Wakehama 18.4 101 S M A - - C 6) Tagami, Y. : The local sacred places and the distribution of ON1 Yakumo S. Yubigahama 39.4 247 S L A - - C ON2 Kumano S. Omaehama 19.5 118 T M A - - C temples and shrines in and around the northeastern Japan, ON3 Yama S. Omaehama 6.9 149 T S X - - A Bulletin Fac. of Edu., Toyama Univ., No. 59, pp. 51-64,

244

2004 20) Miyagi Branch of Association of Shinto Shrines : Shrine 7) Urasaki, S. : Relationship between geomorphic character- Directory books of Iwate Prefecture, Miyagi Branch of istics and space composition of ‘Kyu-Kampeitaisha’ Association of Shinto Shrines,1976. shrines in Nara prefecture, J. JILA, No. 72, Vol. 5, pp. 21) Mitsuhashi, T. and Shirayama, Y. : Nihon Kamisama Jiten 787-792, 2009. (Dictionary of Japanese Deities), Daihourinkaku, 2005. 8) Urasaki, S. : Naraken no Mikumari Jinja to Takagama Jinja 22) Takada et al. 13) op. cit. niokeru Kankyo Design Joron (Introduction of environ- 23) Matsuura 17) op. cit. mental design based on Mikumari Shrine and Takagama 24) Takase et al. 14) op. cit. Shrine in Nara prefecture), Art Culture Res. Osaka Univ. 25) Haraguchi, T. : Inundation Areas of 2011 Tsunami in East Arts Grad. School, Vol. 10, pp. 321-340, 2006. Japan, http://www.jsgi-map.org/tsunami (accessed on May 9) Koresawa, N. and Horikoshi, T. : Sacred places and their 15th, 2014), 2011. natural environment as an index for landscape preserva- 26) Urasaki 7) op. cit. tion : A case study on Shinto Shrines located along the 27) Takada et al. 13) op. cit. Hanaore Fault in Kyoto, City Plan. Rev. Special Is. Papers 28) Iwate Branch of Association of Shinto Shrines 19) op.cit. City Plan. CPIJ, No. 39, pp. 145-150, 2004. 29) Miyagi Branch of Association of Shinto Shrines 20) op.cit. 10) Hattori, S. and Nii, A. : Topographical characteristics of 30) Ishii, M., Com. of Edit. : Utatsu Cyoushi (The History of location of sparse village at an alluvial fan, J. JSCE D1, Utatsu Town) , Utatsu Town, 1986. Vol. 68, No. 1, pp. 35-44, 2012. 31) Sawauchi, Y., Suzuki, A. and Nakajima, T. : Miyako no 11) Imamura, F. : Daishinsai no Jittai to Kyokun no Seiri ni Ayumi (History of Miyako), Miyako City, 1962. Mukete (Toward organizing the actual situation and les- 32) Com. of Edit. : Rikuzen-Takata Shishi (The History of sons on the great earthquake disaster), Papers on Earth- Rikuzen-Takata City), Vol. 7, Rikuzen-Takata City, 1998. quake Regeneration Committee, Miyagi Prefecture, 33) Com. of Edit. : Kesennuma Shishi (The History of Kesen- http://www. pref.miyagi.jp/uploaded/attachment/36560.pdf numa City), Vol. 7, Kesennuma City, 1994. (accessed on May 15th, 2015), 2011. 34) Com. of Edit. : Motoyoshi Gunshi (The History of Mo- 12) Miyasaka, T., Nakai, Y. and Osaki, S. : Bichikei to Suigai toyoshi County), Motoyoshi-Gun Soncyo-Kai (Com. of ni Chakumoku shita Sennan Heiya no Jinja Ricchi Tokusei Mayers in Motoyoshi County), 1949. (Shrines location characteristics in Sennan plain from the 35) Section of Edit. : Shizugawa Choushi (The History of Shi- view point of microtopography and water disasters), Proc. zugawa Town), Shizugawa Town, 1991. of 8th LSD Committee, JSCE, No. 8, pp. 235-240, 2012. 36) Board of Edu., Com. of Cultural Heritage Preservation : 13) Takada, T., Umetsu, K. and Kuwako, T. : A study on the Asahigaura-no-Ki (Note of Asahigaura), Board of Edu., deity and spatial arrangement of shrines in tsunami disaster Com. of Cultural Heritage Preservation, 1983. caused by the Tohoku Earthquake, J. JSCE F6, Vol. 68, 37) Com. of Edit. : Kamaishi Shishi (The History of Kamaishi No. 2, pp. 167-174, 2012. City), Kamaishi City, 1977. 14) Takase, H., Yoshida, K. and Kumagai, W. : Jinja wa Kei- 38) Com. of Edit. : Kahoku Cyoushi (The History of Kahoku koku suru (Shrine warns), Kodansya, 2012. Town), Last Vol., Kahoku Town, 1979. 15) Weather News Corp. : Naze Tsunami kara Jinja wa No- 39) Com. of Edit. : Ofunato Shishi (The History of Ofunato kottanoka? (Why shrines survived the tsunami?), Sora, pp. City), Vol. 5, Ofunato City, 1982. 48-55, 2014. 40) Miyagi Branch of Association of Shinto Shrines : Shrine 16) Nikkei Inc. : Report from Dento Bunka (Traditional Cul- Search System, http://www.miyagi-jinjacho.or.jp/jinja- ture) Seminar conducted by Association of Shinto Shrines, search/detail.php?code=310030770 (accessed on 2015.4. Nikkei News Paper, pp. 40, Dec. 16th, 2011. 10) 17) Matsuura, S. : The mega-tsunami along the Pacific Coast 41) Sasaki, H., Com. of Edit. : Motoyoshi Choushi (The His- of Eastern Japan in March, 2011 and the Shrine, Suiri tory of Motoyoshi Town), Motoyoshi Town, 1982. Kagaku (J. Hydr. Sci.), The Forest Science Institute, No. 42) Noda Town : Noda Sonshi (The History of Noda Village), 332, pp. 168-186. 2013. Noda Town, 1992. 18) Endo, K. and Mazereeuw, M. : Historical study on geo- 43) Okada, S. : Nihon Shinto-Shi (The History of Shinto graphical features of shrines in town of Minami Sanriku, Shrine in Japan), Yoshikawa-Kobunkan, pp. 31-37, 2010. Miyagi prefecture, J. JILA, Vol. 78, No. 5, pp. 693-696, 44) Urban Planning Section, Chief Cabinet of Minister of 2015. Home Affairs 1) op. cit. 19) Iwate Branch of Association of Shinto Shrines : Shrine Directory books of Iwate Prefecture, Iwate Branch of As- (Received October 5, 2016) sociation of Shinto Shrines, 1988.

245