Pathway to Japanese Literature

An Unexpected Snowbound Journey

Just as I was thinking of continuing the Dewa region survey I had begun last November during this spring’s vacation, Administrative Official Ariyoshi contacted me, suggesting I attend the opening ceremony of Yamagata Prefecture’s Committee for the Investigation of Historic Sites, Places of Scenic Beauty, and Natural Monuments.
Seizing this opportunity, I consulted Mr. Abe Masami of the Shōnai region—who had previously assisted me—about investigating local ruins, but he replied that the snow was still too deep this season for such work.
This is no joke—here, the cherry blossoms are already beginning to bloom.
Can conditions truly differ so drastically even within Honshu when it comes to the Ōu region? Perhaps I should prioritize seeing this snowy spring landscape over investigating ruins.
In that case, the Senboku area of Ugo Province—where the snow likely runs deeper than in the Shōnai region—would be preferable.
It would be fascinating to tour through the snow-covered ruins of Kanazawa Stockade—a site tied to the Later Three Years’ War that I’ve long wished to see.
I also wished to visit Mr. Shisuisō Fukazawa Taichi, whom I had not met in quite some time.
Mr. Fukazawa had resigned last year from his post as chief of Tango Kuma District and retired to this Senboku region, where he devoted himself to investigating local antiquities—a pursuit dear to him—and now served on Akita Prefecture’s Historic Sites Investigation Committee, such was his zeal that he had recently contributed a mimeographed copy of *A Study of Oga Castle Ruins*.
If the snow rendered field investigation impossible, then even meeting him and hearing his accounts alone would undoubtedly prove valuable.
After returning home last year, Mr. Fukazawa had fallen victim to a devastating fire, in which many historical materials he had painstakingly collected over years were lost.
I also wished to personally offer my condolences regarding this misfortune.
While deciding to extend my journey to Ugo this time, when I promptly inquired about his availability, I was told, "By all means, come."
The arrangements took time, but I finally departed on the night express leaving Tokyo’s Ueno Station on the 27th.

Unlike the Tōkaidō Line, the Ōu Line’s second-class cars were spacious.
Without needing a sleeping berth, he stretched out comfortably, fell asleep around Ōmiya Station, passed through Shirakawa and Fukushima in his dreams, and awoke only after they had already crossed Itaya Pass—the boundary between Ōshū and Ushū.
Looking out the window—though dawn had not yet broken—the snow did indeed appear deep.
This truly was Dewa, he thought.
In areas like Yonezawa and Aka-yu, every open stretch of flatland as far as the eye could see was a vast expanse of snow.
Strangely enough, it had almost entirely disappeared as they approached Yamagata.
Alighting at Yamagata, he called Administrative Official Ariyoshi to confirm his itinerary, then transferred to the next train and continued northward.
North of Tendō, Kannomachi, and Tateoka, the snow gradually increased again.
On flat ground, the snow measured about two shaku (60 cm) deep; where it had been shoveled up along both sides of the tracks, there were mounds reaching heights of around eight shaku (2.4 meters).
Snow scraped from the ridges of private homes’ roofs formed levees encircling the houses; in some places, the entrances dipped low into sunken chambers.
This way—as one could now see—they effectively blocked the bitter winds and made winters passable in relative warmth.
Along the railway tracks lay a cemetery.
The stone monuments were buried deep beneath the snow.
In that place stood two newly interred graves erected forlornly within hollowed-out depressions resembling mortar bowls where snow had been shoveled away—a sight all the more poignant.
In snow country, even funerals are no simple matter.

North of Shinjō—around Kamabuchi and Oyo—the snow grew ever deeper as we approached the foothills, while here and there along the mountain slopes lay traces of avalanches.
In front of the station, reed screen fences had been constructed as blizzard barriers.

A fellow passenger who appeared to be a civil engineering contractor said:
“I’ve been involved with railway construction since they first built this Ōu Line, but when it comes to snow here—well, folks from other regions couldn’t begin to imagine it.”
“Look at that mess out there! No matter what precautions you take midwinter, you’ll never beat nature’s raw power.”
“And those diet members and newspaper hacks? Won’t even bother coming to see for themselves! Every time tracks get blocked by snowstorms—‘Preventive measures are insufficient!’ ‘Snow falls every year without fail!’—they just carp and criticize.”
“Sure—if we built reinforced concrete avalanche barriers or brick snow sheds like tunnels everywhere it snows heavy? Maybe we’d hold out against any blizzard! But keeping that up year-round? Prohibitively expensive! And downright ugly when there’s no snow!”

Indeed—that rang true enough.
Humanity might inch toward conquering nature’s forces someday—but not yet fully achieved today’s age.
After all—in former times lacking such transport systems—people likely accepted occasional rail stoppages from heavy snow like travelers once did Ōi River ferry suspensions during floods.
Yet these modern “enlightened” souls found such resignation intolerable now—
An era where laborers—who’d never dreamed themselves contenders—now banded together to challenge capitalists head-on!

Box Snow Vehicles and Monpe

Alighting at Gosannen Station just before twelve o'clock, I found Mr. Fukazawa waiting at the ticket gate.
Even the former district chief cut a figure in his hometown wearing a bird-hunting cap and monpe work pants.
Monpe—a hybrid between hakama trousers and Western-style pants—proved indispensable in snow country.
Their construction varied slightly by region: Tachitsuke (abbreviated to Takke), Sarubakama ("monkey hakama"), Fumikomi ("step-in"), while in Shōnai they were called Matashari.
Originally worn by hunters and mountain folk, this garment remained widely used across snow country even today.

Accompanied by Mr. Fukazawa, he took a brief rest at an office of a certain company in front of the station before being guided—together with Shrine Priest Miura Kenro of Kanazawa Hachimangu Shrine—to the residence of Mr. Ebata Shinnosuke in Nakajima, Iizume Village, where they all found lodging for the night.
From the station to Mr. Ebata’s residence—approximately twenty-five chō (2.7 km)—he was transported for the first time in his life by what they called a box snow vehicle.
No—not only had he never ridden in one, but this was also his first time seeing such a vehicle.
On snowy surfaces, rickshaws proved utterly ineffective, so rickshaw pullers temporarily switched to operating snow vehicles during the winter months.
A box just large enough to seat one person was mounted on a sled and covered with a cloth canopy; with a thick floor cushion laid out and wrapped in blankets like a Daruma doll while clutching a brazier, it proved to be an excellent mode of transportation that kept the cold at bay.
They pushed it from behind like one would a baby carriage.
They attached a rope and pulled from the front.
When snow is abundant—particularly during icy periods—it is said to move quite nimbly, but now, during the snowmelt season, this did not seem to be the case.

This winter saw particularly heavy snowfall, and in this Senboku area, it was said that at its peak, eight shaku (2.4 meters) accumulated on flat ground.
Even now, the snow still lay about three shaku (90 cm) deep.
Amidst the vast expanse of snowfield, natural pathways had formed.
Though the surface remained packed down, as meltwater gradually rose from below, even seasoned snow vehicle pullers would occasionally sink knee-deep into the snow.
Along streams and pools, the snow had collapsed here and there, exposing precipitous drops.
On an earthen bridge three to four shaku (90-120 cm) wide, snow piled two to three shaku (60-90 cm) thick slumped from both sides, creating a trapezoidal path barely two shaku (60 cm) across at the top—a path the puller navigated with practiced skill.
Should any part of that half-melted snow give way, the box snow vehicle would instantly plunge into the river below—a possibility that sent intermittent jolts of alarm through us.
On the twenty-ninth, Shrine Priest Miura departed early for Kanazawa while I, accompanied by Mr. Fukazawa, rode the box snow vehicle once more to survey the ruins of Kanazawa Stockade—a site from the Later Three Years’ War.
The distance spanned approximately one ri (3.9 km).
A thick spring haze limited visibility beyond a few hundred meters, rendering every visible inch of the snow-blanketed countryside a pristine white expanse that might rival Siberia’s frozen plains—through this void our sled drifted faintly, weaving an uncertain path.
For me, seeing this snow country landscape for the first time, it struck me with an indescribable strangeness—that this too was part of Japan.
All the more astonishing when considering this was March twenty-ninth; in my native Awa, ordinary cherry blossoms would have long since scattered while double-flowered varieties now bloomed in full splendor.

Beneath our sled lay nothing but rice fields; paying no heed to roads or irrigation channels, they forged paths wherever they pleased across the snow, advancing in whatever direction suited them. Farmers transported fertilizer by sleigh from early morning, each distributing it to what they considered their own land. To an outside observer, it might have seemed worrisome that they could mistakenly leave it on others’ land, but by using distant trees and mountains as landmarks for orientation, they apparently never confused their own plots. In this snow country, people made use of snow-based transportation during that period to move as much as possible in advance, or so it was said. Sleds loaded with timber, sleds loaded with gravel, and others passed by in several long columns. They appeared to be transporting materials for civil engineering works effortlessly while they still could. During heavy snowfall years, it was said that even a single person could transport over a dozen bales of rice with ease. When I had previously imagined from afar, I pitied those in snowbound regions—thinking they must endure months confined to dimly lit houses during that period of immobility. Yet now I realized that very season offered unparalleled transport convenience; even snow, it seemed, had its uses.

Gosannen Station

The railway station where I alighted yesterday was called Gosannen Station. Being newly established in recent years, it did not appear on the Land Survey Department’s maps. Having planned my itinerary for surveying the Kanazawa Stockade site based on maps alone, I had initially been unaware of this station’s existence and consequently purchased a train ticket only as far as the next stop at Iizume Station—until yesterday, when a telegram from Mr. Fukazawa reading “28 Higo 3 Nen Eki Matsu” arrived at my Tokyo residence. Interpreting this to mean I was to wait at Nen Station on the 28th at 3 p.m., I thought What an unusual station name! and searched a map, but found nothing matching it. Only after consulting a travel guide timetable did I first learn of Gosannen Station’s existence.

Gosannen Station was a cleverly coined name indeed.
According to what I had heard, this station stood on common land shared by three villages—Kanazawa Town, Iizume Village, and Kanazawa Nishine Village—and since none could agree to name it after any single village, after much wrangling they had ultimately adopted the name of the Later Three Years' War campaign, as it served as the nearest station to the Kanazawa Stockade site connected to that conflict's battlefields.
Yet the homophonic resemblance between "campaign" and "station" proved undeniably confusing.

From the station to Kanazawa Stockade site—the final relic of the campaign—was approximately one ri (about four kilometers). Unfortunately, a cluster of hills blocked distant views—a regrettable irony given the station’s name. These hills stood completely isolated from other mountains in the flatland, their southern end apparently called Kyōzuka-yama. On its upper part, terraced land encircled the mountain like a headband. It was said this upper terrace might be an ancient burial mound—or perhaps the remains of a moat dug for a fortress. Due to deep snow and insufficient time, we had been unable to conduct a climbing survey. We would need to postpone this for another day. Most land in this area was owned by Mr. Ebata, who I learned planned to soon offer it for creating a large public park.

Mr. Ebata Shinnosuke

Mr. Ebata Shinnosuke, who had been guided by Mr. Fukazawa, was said to rank among Akita Prefecture’s foremost taxpayers, his mansion—rebuilt after a fire some years prior—standing as testament to this status.
The structure was exceedingly grand, fronted by an expansive garden.
Mr. Ebata had long cultivated interests in both gardens and reading; his estate’s woodland stream garden spanned a river channel, its design revealing considerable ingenuity.
Yet with snow still burying the grounds deeply—bamboo snow barriers protruding like spears to protect trees—the terrain remained wholly obscured, permitting us only to grasp its scale.
His ambition to convert the hills around the station into a park undoubtedly sprang from these passions.
His book collection already numbered substantially, and he continued avidly acquiring new publications with evident intent to eventually open a public library.
That he allocated a storehouse considerably removed from the main residence for this purpose struck me as prudent.
Every collector must exercise such foresight.
To amass antiquities and rare texts only to lose them all in a fire constituted nothing less than transgression.
Thus Mr. Fukazawa—who had incurred this transgression the previous year and now lacked reference materials—applied himself diligently to organizing the Ebata library’s holdings, compiling catalogs while conducting repeated studies.
Any volumes Mr. Fukazawa required were furnished through the Ebata collection—an admirable resource now at his disposal.

Guided through what was nominally part of the same estate grounds yet required scaling snowy peaks and descending into valleys, they finally reached the library.
The library was a still-new storehouse-style building with a tatami room attached to one side.
Mr. Fukazawa was there engaged in organizing books and perusing them.
The library catalog had yet to be properly organized, making it inconvenient to peruse, but the collection spanned such diverse fields that this repository seemed to lack any distinctive character.
Mr. Ebata too was troubled over how best to develop this collection in the future.
Thereupon, I informed them how my hometown’s Mr. Ishihara’s Gokō Library had confined itself primarily to history and geography, with particular focus on collecting materials related to Awa, and recommended that—given Mr. Fukazawa’s involvement as a history enthusiast—it would be advisable to develop the collection along these lines, striving with the ambition that this library should become the place where one could find anything related to Dewa’s historical materials.
Mr. Ebata also agreed to this.
I had no doubt that Mr. Fukazawa shared this view.

I borrowed *Masumi Yūranki* from the collection and skimmed through parts here and there.
What I had included in the materials section of the previous issue was a passage from it.

On the night of the 28th, I was kindly lodged at Mr. Ebata’s residence together with Mr. Fukazawa and Mr. Miura.
He had stone tools and pottery—reported to have been excavated from the nearby Ishina Hall—sent over and showed them to us.
The pottery was naturally of Ainu-style, and while most stone tools were common Ōu region types, one rare specimen—an ito-nade resembling a hemp-peeling tool—stood out.
After all, this was the Senboku region.
Ainu ruins seemed to be numerous.

Tenant Farmers and Leather Workers

The night at Mr. Ebata’s residence was spent in beneficial conversation, unaware of the deepening hour.
Among these discussions, what I found particularly interesting was learning that in this region, tenant farmers are referred to as taya.
Taya of course derives from the term "field hut".
In other regions, women’s menstruation is generally called taya, and the period of menstruation is referred to as “being in taya”.
Regarding taya, I introduced Mr. Hashikawa Tadashi’s “A Study of Taya” in Volume 4, Number 4 of Minzoku to Rekishi, but that primarily concerned the Honganji’s taya.
The practice of calling menstruation taya undoubtedly derived its name from our nation’s customary aversion to blood impurity—a tradition that forbade cohabitation with menstruating women within residences and particularly barred them from approaching household altars. Just as fishing villages typically established monthly huts along coastlines where women spent their cycles, rural communities utilized readily available field huts for this seclusion period.
A taya (field hut), as its name suggests, was a small structure erected in rice fields where farmers who normally resided in village homes would lodge there during busy agricultural seasons to avoid the inconvenience of commuting morning and evening to their fields.
Of course, fields near villages likely did not require this practice, but even so, huts must have been necessary for storing harvested rice or threshing grain—and they likely repurposed these structures as the aforementioned isolation quarters for menstruation.

Now, the fact that in this Senboku region specifically, tenant farmers—referred to elsewhere as "half-people" or "intermediaries"—are called taya warrants consideration.
Historically, the Ōu region was originally land of the Emishi, and particularly within Dewa’s Senboku three districts, remnants of Emishi captives and prisoners persisted relatively late. Thus, as seen in the provisions of the Taihō Code, residences were always protected within castle walls, and the custom of only venturing out to dwell at cultivated fields during farming seasons is thought to have continued until comparatively recent times.
These dwellings at the cultivated fields were none other than the taya.
Even after the need for defense against the Emishi had passed, the landowning farmers continued residing in their original villages away from the cultivated fields, while the taya at the cultivated fields came to be occupied by so-called serfs—those who owned no land but were bound to till it—wouldn’t this have given rise to such a custom?
Serfs were precisely the category of later tenant farmers.
Thus, there arose the appellation 'taya' for tenant farmers.

In Senboku as well as in the Shōnai region, it was said that those who handled leather—a former special class—were called leather workers.
As I had written in the “Shōnai Miscellaneous Matters” section of this journal’s January issue, this was undoubtedly an abbreviation of Hakuraku (Master Herdsman), referring to veterinarians for cattle and horses.
On January 28th of this year during a lecture on discrimination abolition in Yamato Goshomachi, I borrowed community-related documents from Mr. Maekawa Masukichi of Nishimatsumoto Village—among them lay a scroll titled *Hakuraku densho* illustrating bovine disease treatments.
In the Kinai region too these people conducted veterinary care for livestock—hence their designation as Hakuraku.
This Senboku region had originally contained small numbers of leather workers who engaged in hide crafts and performed manzai routines; some descendants reportedly still exist today stretching drum and shamisen skins.
However most had now shifted occupations—their residences relatively unrestricted—with occasional intermarriage occurring and no significant discrimination endured.
This was precisely how matters ought to be.

Protective Walls and Standing Screens

In areas with deep snow, if roof snow was not regularly cleared away, there was danger that flimsy houses might collapse under its weight.
Yet scraping it off caused accumulations around dwellings that rose to eclipse eaves.
Even without such efforts during winters of heavy snowfall like this one, drifts reaching eaves on low houses were hardly uncommon.
Thus in these regions people erected *yukigakoi* snow barriers around homes to prevent drifts from pressing against walls.
Snow country houses typically enclosed three sides with walls—an entrance on one face and sparse small windows elsewhere—so residents usually raised wooden posts along walls to bind reed screens or rush mats into surrounding barriers. However, at residences like Mr. Ebata’s with urban-style formal rooms, they had prepared board-made shutters resembling traditional lattice panels beforehand and erected these externally.
All served as defenses against direct snow contact with walls.
In the *Wamyōshō*,

Bound Wall: The *Shiming* states, "A bound wall uses mats bound to the wall." The *Kango Shō* states, "Protective Wall (tatukomo)."

It states this. And it lists them among barrier implements. I had briefly addressed this matter in Volume 6, Number 5 of *Minzoku to Rekishi* while discussing how beggars were called "*okomo*," but now, observing rush screens and similar materials erected around houses in this snow country as protective wall installations, I finally understood why "bound walls" referred to objects affixed to walls with mats, and why "protective wall" came to be written with those characters yet read as *tatukomo*. Of course, these are installations used only during snowy seasons here in snow country, but I believe that in the crude houses of antiquity, people similarly erected rush screens as protective walls against ordinary wind and rain erosion. The *Shiming*'s explanation of "bound walls" perfectly illustrates how *tatukomo* functioned in this application. Moreover, that [the text] included this among barrier implements in the *Wamyōshō*, listing it alongside curtains, folding screens, and reed blinds, shows how later generations preserved archaic terminology even as they repurposed these items—thus we should interpret the earlier citations from *Shiming* and *Kango Shō* not as describing barrier implements per se, but rather as explanations for applications belonging to the category of wall-building tools.

Snow-Covered Kanazawa Stockade Site

On March 29th, relying on that familiar box snow vehicle from Mr. Ebata’s residence, I was conveyed across roughly one ri of snow-covered countryside to Kanazawa Town.
On flat ground, it glided freely and proved exhilarating; however, traveling over thawing snow—with streams flowing beneath and occasional collapses into water pools—left us anxious at times.
Once we entered a village, it being this transitional season between snow paths and bare earth, the snow scraped from rooftops still stood piled into steep mountains here while elsewhere had been cleared to form deep valleys. Each time we were jolted up and down through these undulations, I thought how arduous this snow vehicle journey truly was.

The site of Kanazawa Stockade, where Kiyohara no Takehira and Ieharu made their final stand, stood on a hill in Kanazawa Honmachi. The river flowing at its base was called Kuriyagawa, and crossing it brought one directly to the stockade’s approach path. Let me clarify—this Kuriyagawa was entirely distinct from the Ōshū Kuriyagawa associated with Abe no Sadato’s demise. At the stockade site, Hachiman Shrine had been established, its administrative office positioned at the entrance. Already waiting at this office were Shrine Priest Miura—dispatched earlier that morning from Mr. Ebata’s residence—alongside former Diet member Mr. Itō Naozumi, Kanazawa Town Assistant Mayor Mr. Itō Naonosuke, and Mr. Ebisuya Kamekichi, owner of the local confectionery shop Emisuiya.

Mr. Itō Naozumi was a devoted Kanazawa scholar who had amassed historical materials related to Kanazawa Stockade and the Later Three Years’ War, compiling them into an extensive series titled *Kanazawa Shisō*. When Mr. Fukazawa’s home burned down the previous year, he had written a memoir called *A Record of Encountering Fire* as a keepsake, distributing mimeographed copies to acquaintances and friends—yet on the very day Mr. Itō received and read this account, his own residence too had coincidentally caught fire. As Mr. Fukazawa’s article containing these reflections had reached me at that time with Mr. Itō’s essay appended, I had already known of Mr. Itō’s name and his reputation as a man of singular dedication beforehand—yet meeting him now stirred an unexpected sense of familiarity, as though greeting an old friend.

Ebisuya-san stood as Mr. Itō’s equal in Kanazawa expertise—a master of polychrome painting who had produced illustrated albums cataloging shrine treasures and collaborated with Mr. Itō to create over a dozen supplementary volumes restoring gaps in the *Later Three Years’ War Picture Scrolls*, all through meticulous research. Such fervor lay far beyond ordinary capability.

From the shrine office to the stockade site with its shrine hall—a distance of about six chō—the snow lay so deep that ascending in traditional Japanese attire and wooden clogs seemed utterly impossible. Borrowing the usual monpe trousers and changing out of my hakama, I donned borrowed rubber boots over my tabi socks and carefully picked my way up the already packed snow. In areas without foot traffic where rubber boots alone proved inadequate, I strapped snowshoes over my boots and followed Mr. Itō’s lead, steadily trampling the thawing snow from below. Occasionally breaking through the surface, I would sink up to my thighs even with snowshoes still strapped on, and extracting them proved no easy task.

The hilltop had been leveled into two tiers, with the hall of Hachiman Shrine standing on the upper level. A plaque bore the inscription "Seihachiman Shrine." Moreover, with structures such as secondary enclosures and western enclosures among others, its layout differed from that of an ordinary mountain fortress and proved remarkably intricate. At what was said to be the provisions storehouse site, we found the usual carbonized rice husks. Though commonly called "burned rice," this did not necessarily mean it had been intentionally charred. Among items I viewed in Mr. Ebata's collection the previous night were specimens that had flowed down Kuriyagawa River in clumps—a curious phenomenon suggesting large quantities of provisions had been carbonized and buried intact here.

During the Later Three Years' War, when Kiyohara no Takehira and Ieharu made their final stand at this stockade and tenaciously defended it, even the renowned Hachiman Tarō found himself confounded with no means of breaching its defenses. That he ultimately resorted to siege tactics stemmed not only from Takehira's fierce valor but more fundamentally from this stronghold's impregnable fortifications. Records claim even Takehira's forces were overwhelmed when provisions ran out, yet the puzzling abundance of so-called burned rice still unearthed today suggests another possibility—that this castle may later have been garrisoned by other warlord clans.

While historical accounts mention occupation by the Kanazawa clan, we cannot definitively rule out others having held it subsequently. Thus the present stockade site likely preserves not its original Later Three Years' War configuration but rather reflects numerous later modifications.

Mr. Katō pointed out each nearby feature in detail and provided explanations. To the south lay Hirumo Swamp with its emerald waters—the very spot where Takehira had hidden himself by draping algae over his head; nearby rose Jin’gaoka Hill, site of Yoshiie’s encampment; westward stretched the low slopes of Jinkan Hill, location of the main castle’s outpost; and there, the deep trench that marked the remains of their siege moat—each feature was delineated with the precision of one tracing lines upon a palm. On the return train journey, I found myself sharing the carriage with a group of army officers. From their conversation, it emerged that military exercises were scheduled near Kanazawa Town the day after tomorrow, and they were to receive a lecture from Mr. Katō Naozumi. Mr. Katō was truly an indispensable figure when it came to the Later Three Years’ War and Kanazawa Stockade. Descending the original path, I rested at the shrine office and viewed the displayed treasures along with a specially prepared copy of the Later Three Years’ War Picture Scrolls. Then, escorted once more by snow vehicle, I headed toward Gosannen Station.

Kanazawa Hachimangu Shrine's All-Night Vigil

The Hachiman Shrine at the Kanazawa Stockade site was said to have been established by Minamoto no Yoshiie, who enshrined a divided spirit from Iwashimizu for the pacification of Dewa Province.
A ridgepole plaque from Satake Yoshinobu's shrine renovation in Keichō 9 (1604) bore the inscription "Guardian Deity of the Six Districts of Dewa Province."
Yet in later eras, only Kanazawa district remained as its parishioners.
As Hachiman Shrine—whose head shrine at Iwashimizu had since ancient times conducted hōjō-e life-release ceremonies—particularly abhorred bloodshed and maintained strict taboos against ritual pollution, killing within its precincts remained prohibited even now. During its festival months of January and August, from the new moon's first day through the fifteenth day, consuming fowl or game meat was forbidden; should any parishioner die during this period, they were not interred within the village but instead borne to neighboring settlements' boundaries for funerals.

At Hachiman Shrine’s hall, an event called “o-tsuya” (all-night vigil) was held.
Men and women gathered to hold song duels and determine the victors.
The songs followed fixed forms, but the lyrics were often improvised and sung to a melody called *Nigata-bushi*.
Those who ran out of lyrics or whose voices gave out were declared the losers and withdrew, making way for new challengers.
It typically began around midnight and continued through the night.
Kakeuta likely referred to singing verses that took cues from—and thematically connected to—either the wording or content of one’s opponent’s preceding song.
The term *utagaki* in ancient times likely denoted gatherings where participants lined up to form crowds and exchanged songs; this *kakeuta* represented another variety of such communal singing.
The Utagaki Mountain mentioned in the *Settsu Fudoki*, the Kagai of Tsukuba found in the *Man’yōshū* and *Hitachi Fudoki*—all were likely variations of the same tradition.
In my hometown as well, during summer nights, young men and women often gathered in temple courtyards and such places, mimicking wheat-pounding motions while improvising new songs that played off their opponent’s lyrics—events where they competed to determine a winner were frequently held.
They called it Osugata.
This was because they began by singing many songs that praised or teased their opponent’s appearance.
“If we were to behold your noble figure from here, you’d look just like Yamada’s boar-scaring figure”—that sort of thing.
In ancient times, such practices were likely carried out throughout various regions.
And indeed, these practices were still performed during the annual festival at Kanazawa Hachimangu Shrine.
In Mr. Itō’s *Brief Record of Hachiman Shrine*, it stated: “The village north of Hachiman Shrine was called Kitaura of Semboku, and in this region, many young women of marriageable age would not wed unless they visited the shrine during the annual festival and performed an overnight vigil.”
“They call this o-tsuya (all-night vigil), and it is said to still be practiced today.”
Just like the *utagaki* song duels of Tsukuba, this too likely originated as an opportunity for unmarried women to choose their husbands.

Semboku’s Emishi Captives

The name Semboku now refers to a single district situated between Hirataka and Kawabe Districts, but it originally signified a broad area north of the mountains—a general designation for the region centered on the plains along the upper Omotegawa River, north of the mountain range marking Uzen’s border. The present-day three districts of Ogachi, Hirataka, and Semboku all fall within what was historically called Semboku. This was at times collectively referred to as Semboku District. The current Semboku District roughly corresponds to the area designated as Yamamoto District in the Engishiki, constituting merely a portion of historical Semboku. As for present-day Yamamoto District, during the Engishiki era it remained unincorporated barbarian territory—the very location where Nushiro District had existed during Abe no Hirafu’s expedition. Yet during the district name reorganization in the fourth year of Kanbun (1664) under Tokugawa rule, authorities gravely misapplied these historical realities: mistaking Semboku for a single district’s name, they attached it to what was originally Yamamoto District while reassigning the old Yamamoto designation to a previously unnamed area.

Semboku is also found written as Senpuku in the *Azuma Kagami*.
Due to the characteristic Tohoku dialect making "Sempoku" and "Senpuku" indistinguishable phonetically, Kamakura-period warriors arbitrarily transcribed what they heard from local inhabitants using whichever characters seemed fitting.
This region—situated along upper river reaches and bounded by southern mountains—experienced relatively delayed exposure to Japanese cultural influences.
Imperial governance first reached Dewa through coastal areas; by Empress Saimei’s reign, districts had already been established via Abe no Hirafu’s expeditions in regions like Akita and Nushiro (present-day Noshiro), with assimilated Emishi communities existing as far north as Tsugaru.
Yet this mountainous Semboku territory long persisted as a domain of unassimilated Emishi—so-called “raw barbarians”—requiring travelers moving between Mutsu Province’s capital at Taga Castle and Dewa’s Akita Castle to first descend westward along the Mogami River before detouring through Atsumi District on the western coast.
Deeming this circuitous route impractical, authorities during the Tenpyō era (729-749 CE) subjugated the Emishi in this area, opening direct access from Mogami District to the upper Omote River via the Ogachi route.
Thereafter began gradual transfers of mainland Japanese settlers to Ogachi Castle alongside advancing colonization of the Semboku Plain.
Though Emishi increasingly assimilated into mainland society as “captives,” the region long remained a contested borderland where ethnic Japanese and indigenous populations intermingled.
In Gangyō 4 (880 CE)—approximately 140 years after Tenpyō-era expansion—the governor of Dewa Province submitted a petition stating as follows.

Among the governed districts, the three northern mountain districts of Ogachi, Hirataka, and Yamamoto lay distant from the provincial capital and bordered rebel territories.
Since ancient times, rebel barbarian tribes had lived intermingled with the people, exploiting every opportunity to become a festering affliction.
In recent years they had suffered frequent poor harvests and found themselves distressed amidst famine.
Should compassionate relief not be granted, harmony between people and Emishi would prove difficult; we humbly entreat that taxes and corvée be suspended for two years to grant respite to our exhausted populace.

In response to this petition, an imperial decree granted a one-year tax exemption and distributed 6,200 koku of emergency grain to the 803 Di captives across three districts.
The term “Di captives” (狄俘) equates to “barbarian captives” (夷俘)—both denoting unassimilated Emishi tribesmen.
In Japan’s case, though these Emishi shared Ainu ethnicity, those in Echigo-Dewa regions came to be called “Emishi Di” or “Di captives” by adopting Chinese directional terminology—designating Ōshū as East and Echigo-Dewa as North.
During Gangyō era, Semboku still harbored numerous unassimilated tribesmen.
From Engi era onward, provincial administration fell into severe disarray.
In Ōshū, resurgent Emishi clans absorbed even established districts back into their territories.
Thus Abe no Yoritoki—leader of Emishi captives—seized six central Rikuchū districts, defying provincial authority and sparking the Former Nine Years’ War.
At this juncture, Semboku appears to have been occupied by fellow Emishi chieftains—the Kiyohara clan.
Minamoto no Yoriyoshi, Mutsu Governor who had subdued Abe forces, and his son Yoshiie—renowned capital warriors—proved unable to conquer Abe no Sadato swiftly.
After twelve years total, they finally destroyed him by enlisting aid from Kiyohara chieftains of Semboku captives.
Yoriyoshi’s persuasion tactics—marked by extreme humility and deference—become clear through how Kiyohara later treated him as kin.
Thus Kiyohara brothers Mitsuyori and Takanori, leading over ten thousand cavalry with clansman Yoshihiko Hidetake, annihilated Abe forces.
Takanori earned appointment as Defense Headquarters Commander-in-Chief for his merits; having governed Semboku before relocating to Izawa to project power across Ōshū, their influence undoubtedly eclipsed Abe’s.
This inexorably led to the Later Three Years’ War’s outbreak.
Takanori stood as first Emishi-born Commander-in-Chief.
Clansman Hidetake’s surname Yoshihiko—equivalent to Yoshiyako—remained common among captive lineages.
At Kanazawa’s fall during Later Three Years’ War, Yoshiie reviled Kiyohara clan, declaring Takehira—a “base barbarian”—owed his undeserved commandership solely to Yoriyoshi’s patronage.
Emishi domination across Mutsu-Dewa would mean tolerating rivals’ “snoring beside one’s pillow”—an intolerable scenario for Yoshiie’s faction.
Whatever proximate causes sparked Later Three Years’ War, its roots lay in unavoidable civilized-barbarian conflict.
Even Yoshiie couldn’t have overcome Kiyohara relying solely on his forces.

He ultimately succeeded in destroying them by relying on the aid of Fujiwara no Kiyohira—an Emishi captive—and thus, after Yoshiie’s term had concluded and he returned to the capital, the credit fell to Kiyohira, who became Commander-in-Chief of the Defense Headquarters in place of the Kiyohara clan and subsequently extended his influence over both Mutsu and Dewa provinces.
Later, Minamoto no Yoritomo’s Ōshū Campaign in the fifth year of Bunji (1189), though ostensibly justified by the Fujiwara clan’s harboring of his brother Yoshitsune, was at its root an inevitable clash between the civilized and barbarian peoples that could not be avoided.
The final demise of the Semboku Emishi captives’ power can indeed be said to have occurred during Yoritomo’s Ōshū Campaign.

Yamagata Taian-ji – The Akimoto Family’s “Purse Temple”

I boarded the 5:00 p.m. train at Goshannen Station and arrived in Yamagata at 9:45.
In that spacious second-class car, there were only two or three passengers.
Having fallen into a pleasant sleep, I remained unaware of our arrival at the station; though fortunately the train terminated at Yamagata, it nearly ended up being shunted into the depot.
Traveling in the same vehicle as Administrative Official Ariyoshi and Mr. Abe (Masami), who had kindly come to meet me at the station, I was escorted to Yamagata Hotel where I would be staying for the night.
Mr. Abe (Masami), serving as a Committee Member for the Investigation of Historic Sites, Places of Scenic Beauty, and Natural Monuments, had arrived early for tomorrow's opening ceremony.
Fortunately sharing lodgings with him, I was delighted to obtain this opportunity to leisurely hear about aspects of the land I had neglected to inquire about during my visit last December.

The hotel was originally the site of the Akimoto family’s Taian-ji Temple, and its garden is said to still retain the original form of its woodland spring garden from that time.
The Akimoto family, as fudai daimyo, experienced repeated domain transfers. In Yamagata, after relocating from Kawagoe in Bushu Province during Meiwa 4 [1767], they were transferred to Tatebayashi in Joshu Province in Kōka 2 [1845], their tenure lasting less than eighty years until they exchanged domains with Mizuno Echizen-no-kami. Yet even during this period, they constructed Taian-ji Temple there as their family mortuary temple.
This temple was colloquially known as the Akimoto family’s “Purse Temple,” said to have been carried along like a waist pouch with each domain transfer.

This I had learned from Mr. Abe; below I noted down two or three interesting points from his accounts.

Shōnai Lion Dance, Kagura, and Saidō

In the south of Matsumine Town, Atsumi District, there was a village called Ōnuma.
There had long existed a dance known as the lion dance there, performed by children from farming families.
They donned lion masks to perform, accompanied by bamboo clapper players, a chanter, and two staff bearers holding poles measuring about five shaku (approximately 1.5 meters).
This dance was reportedly invited to appear not only at their village's tutelary shrine festivals but also at those of neighboring villages.
This was undoubtedly a surviving relic of dengaku theater—akin to Uwajima's Kanako Odori dance described in the March issue or Toyohashi's Oni Festival—practices once widespread across various regions that persisted locally even after their original names faded from memory.

In Honmachi and Shinmachi of Matsumine, there was Kagura. First emerged a performer wearing a tengu mask, holding a trident in hand and wearing tall wooden clogs, executing various movements in time with the flute. Next came the Kagura proper. What appeared at first glance to be an ordinary lion dance—two people manipulating a large lion, one holding its head and another its tail—soon revealed deeper ritual significance.

The Saidō festival honoring the roadside deity Dōsojin was held annually on January 15th atop deep snowdrifts. The town divided into two factions where barefoot young men wearing only tabi socks and headbands carried out large drums from both sides, loaded them onto sleds, and tucked roughly carved drumsticks at their waists before clashing at the central bridge to vie for passage. These collisions grew so fiercely energetic that injuries frequently occurred—all for a deity that required no shrine building, only huts constructed around January 5th where children gathered to beat drums in worship. On the final day, they expelled epidemics by sending the deity’s spirit to the town outskirts through ritual procession. It was said those earlier clashes reenacted historical conflicts from this very expulsion rite.

The above summarizes Mr. Abe's account, though there may be errors in my hearing or transcription. I had made arrangements to trouble the same gentleman for a detailed account by having him write it himself.

Concerning the Investigation of Historic Sites in Yamagata Prefecture

Today finally saw the convening of Yamagata Prefecture's Historic Sites, Places of Scenic Beauty, and Natural Monuments Investigation Committee at the assembly hall.
From the Home Ministry came Dr. Yoshitaka Miyoshi, who offered various cautions regarding investigative methods.
I too participated in this gathering and endeavored to deliver a lecture concerning historic site investigations within Yamagata Prefecture.
I expounded fundamental principles: that historical research must progress through both documentary records and field surveys; that for ancient periods lacking written materials particular emphasis must be placed on onsite investigations; and especially that in the Ōu region—where such materials prove particularly scarce—this deficiency must be remedied through examination of archaeological remains. Having established these points, I proceeded to discuss the historical development of regional settlement in Yamagata.
In broad strokes I outlined: research concerning the Emishi people as primordial inhabitants; the necessity of investigating Stone Age sites as their remnants; how imperial governance extended into Dewa from both Echigo and Mutsu provinces; matters relating to Emishi captives and subjugated populations; investigations of fortress sites and burial mounds; fluctuations in power between civilized and indigenous peoples; and the ultimate fate of the Emishi—before bringing my lecture to conclusion.

After the lecture, I received a visit from Mr. Watanabe Tokutaro and, at his invitation, was treated to lunch at Chitose-tei.
He had served for many years as principal of Yamagata Commercial School while concurrently directing the prefectural library, having accompanied me on my Manchuria-Korea expedition some years prior.
I was delighted to receive two volumes of *Yamagata Prefecture Economic Historical Materials* edited by his esteemed elder brother Mr. Watanabe Shōsaburō.
My social history research would find no small benefit in these materials as well.

That afternoon saw an introductory meeting for the investigation committee members, where I participated and presented my views regarding historic site investigation policies.

Yamagata Castle Site

In the afternoon, I received a guided tour of Yamagata Castle Site and surveyed it.
Prior to the Meiji Restoration, it had merely served as the Mizuno clan’s 50,000-koku castle; however, being originally the Mogami clan’s 570,000-koku fortress, its scale remained extraordinarily vast—something beyond the means of daimyo with 50,000 or even 100,000 koku stipends.
After the Mogami clan’s dispossession in Genna 8 [1622], one could say there was scarcely any fixed castle lord; under the Torii clan and others, the domain changed hands among eleven families over a mere 145 years, during which it also periodically fell under direct shogunate jurisdiction.
Thus when the Akimoto family was transferred there with a 60,000-koku stipend in Meiwa 4 [1767], the castle’s deterioration had become exceedingly severe; however, they restored it and cultivated 3,000 koku of rice fields within the outer moat to serve as their tax base.
Yet when the Mizuno family was transferred there from Hamamatsu with a 50,000-koku stipend in Kōka 2 [1845], even this 3,000-koku area within the castle was counted toward their tax base, effectively confining the complex to within the Ni-no-Maru precincts.
An additional moat had been dug around the central honmaru (main keep), but now that it had become military barracks, this moat was filled in without leaving any trace.
At the officers' assembly hall, I viewed several scrolls of maps showing the castle’s condition from the Mogami period through to before Akimoto renovations.
During the Mogami era, residences for samurai had been allocated not only within Ni-no-Maru but also in the so-called 3,000-koku area beyond—evidence of former prosperity.
Yet when some trivial matter incurred shogunate displeasure leading to clan dispossession, these arrangements promptly scattered into oblivion.

*Obako* Dance

At the evening dinner at Yoshanro, I had the honor of hearing what they call the *Obako-bushi* folk song of Shōnai.
The Obako Dance was explained as a hand dance performed by young maidens.
According to the explanation, “obako” refers to young women—a term deriving from how younger brothers are called “oji” and younger sisters “oba.”
The “ko” in “obako” likely represents the Tōhoku dialect suffix frequently attached to noun endings.
Even in Tokyo areas, people say *sumikko* for corners and often add “-ko” endings like *unko* (feces) or *shikko* (urine), but this occurs particularly often in Tōhoku.
Since cows are called *beko*, they add “ko” to make *bekoko*, with calves consequently called *kokko* from *bekoko no kokko*—following that pattern.

The lyrics of Obako-bushi were an innocent display of rustic charm, brimming with local dialect.
I noted down a small portion of them below.

"I went out to the edge of the rice paddies thinking, 'Is Obako coming?'—Kobaete kobaete—but no Obako came; instead, some useless tobacco peddler (meaning 'or something like that') showed up.
Kobaete, kobaete"

I peeked out the small back window wondering if Obako was there—*Kobaete, kobaete*—but no Obako to be found, just some useless granny (or the like) at a spinning wheel—*Kobaete, kobaete*.

Worried that Obako—who’d been out of sight lately (*meaning she hadn’t been seen around*—might have caught a cold—*Kobaete, kobaete*—but though she wasn’t ill, her parents kept her caged like a bird (*the Tōhoku dialect features frequent voiced consonants*)—*Kobaete, kobaete*.

Obako’s heart fluttered like water pooled on lotus leaves at the pond’s edge—*Kobaete, kobaete*—just a slight touch (*de do* meaning ‘if you touch’) sent her tumbling down in a heap (*so ma* meaning ‘immediately’)—*Kobaete, kobaete*.

Obako had just dozed off (*shitareba*) when a young *kariyume* (hunter) came thrusting out his short spear—*Kobaete, kobaete*—Hunter, what are you doing? Hunting bears is your trade (*shōbae*), you know—*Kobaete, kobaete*.

At Sakata Sannōzan, I wrestled with Ebi-nko (shrimp) and kanjika-ko (kanjika-ko being *kajika* = gudgeon)—*Kobaete, kobaete*—Why’s my waist bent again? Having been thrown by kanjika-ko in sumo—that’s why my waist is bent.
Kobaete, kobaete.

The eel takes a bride—with an eight-eyed go-between (*nakādo*), the loach's child gets wed—*Kobaete, kobaete*—the bride resembles her mother-in-law (*shiyōto*), weak-waisted and meek—whether for better or worse—*Kobaete, kobaete*.

Such was the general gist of it.
The interlude refrain *Kobaete* signified “if [she] comes, it would be good,” being a dialectal expression from the Sakata region.
Even within Shōnai itself, in Tsuruoka this was apparently called *Kobaicha*.

The Haguro Naked Festival

At the dinner gathering, among the various informative accounts I received from the gentlemen present, I noted down one regarding the Haguro Naked Festival. Held around January 31st on New Year's Eve, villagers—divided into two groups—pushed against each other while completely naked amidst the deep midwinter snow to expel pestilence mites, or so it was said. They created a large figure of a pestilence mite from straw, cut it out, and scattered it. The hot-blooded young men vied to seize it. If they carried it beyond a certain boundary, there was no dispute since ownership would then be acknowledged. They would take it back and place it at the entrance of their houses. They burned the remaining portions in a designated area. It was said that whichever side finished burning theirs first became the winner. What I found intriguing was that they called the leaders of both groups in this pushing festival *Hijiri*. "Hijiri" meant 'holy one,' a term commonly referring to Buddhist ascetics, but there they served as ritual leaders conducting the pestilence mite expulsion ceremony. These Hijiri were said to leave their homes for one hundred days, retreat to Haguro Mountain, and undergo ritual purification. As this festival drew a large crowd, they dug through the accumulated snow to create chambers where tea shops were opened, serving pilgrims tea and sweets along with sake and food.

Customs of this kind appeared to exist throughout the Ōu region, and it was said that the Somin Festival in Kuroishi, Esashi District, Rikuchū Province also resembled these practices. In Volume 5, Issue 4 of *Minzoku to Rekishi* (Ethnos and History), an article titled “The Ugo Pushing Festival,” reported by Mr. Hosoya Norisada of Ugo Hiraka District, also resembled these practices. As for the aforementioned Haguro rituals, since I had requested detailed reports from Mr. Katō Masayoshi and Mr. Takahashi Sakae, there would likely be an opportunity to present them in this journal in due course.

That night, I returned to the capital by the last express train.

My stay in Dewa lasted barely three days, but for someone like me who knew so little of the region, the abundance of novel experiences brought genuine delight.
Upon returning, I found Tokyo’s cherry blossoms already in bloom.