Folklore and earthquakes: Native American oral traditions from Cascadia compared with written traditions from Japan (PDF)

Folklore and earthquakes: Native American oral traditions from

Cascadia compared with written traditions from Japan


Department of Earth and Space Sciences, University of Washington, Box 351310,


Department of History and Program in Religious Studies, 108 Weaver Building,

With Contributions from D. CARVER3




Seattle, WA 98195-1310, USA (e-mail:

The Pennsylvania State University, University Park, PA 16802, USA



Carver Geologic, P.O. Box 52, Kodiak, AK 99615, USA


13797 Silven Ave NE Bainbridge Island, WA 98110, USA




FEMA, Federal Regional Center, 130 228th St, SW Bothell, WA 98021-9796, USA


Dept of Anthropology, Douglas College, New Westminster, BC, V3L 5B2, Canada


Department of Anthropology, Room 13–15, Tory Building University

of Alberta, Edmonton, Alberta, T6G 2H4, Canada


Chief Councilor Huu-ay-aht First Nation, P.O. Box 418,

Duwamish Tribe cultural resources expert, Duwanish Tribal Services, 4717 West Marginal

10Snoqualmie Tribe, cultural resources expert, and great-grandson of James Zackuse,

Duwamish Indian Doctor, The Snoqualmie Tribe, P.O. Box 280, Carnation, WA 98014, USA

12Rm 1297, 1873 East Mall, University of British Columbia, Vancouver BC V6T 1Z1, Canada

14Makah Cultural and Research Centre, Makah Tribe, P.O. Box 160 Neah Bay, WA 98357, USA

Port Alberni, B.C., V91 1M7, Canada

Way SW, Seattle, WA 98106, USA

11310 NE 85th St, Seattle, WA 98115, USA

13Simon Fraser University, Burnaby, B.C., Canada

15Olympic National Park, Port Angeles, WA, USA

Abstract: This article examines local myth and folklore related to earthquakes, landslides, and

tsunamis in oral traditions from Cascadia (part of the northern Pacific coast of North America)

and in written traditions from Japan, particularly in the Edo (present-day Tokyo) region. Local folk-
lore corresponds closely to geological evidence and geological events in at least some cases, and the

symbolic language of myth and folklore can be a useful supplement to conventional geological evi-
dence for constructing an accurate historical record of geological activity. At a deep, archetypical

level, Japan, Cascadia, and many of the world’s cultures appear to share similar themes in their con-
ception of earthquakes. Although folklore from Cascadia is fragmentary, and the written record

short, the evolution of Japanese earthquake folklore has been well documented over a long

period of history and illustrates the interaction of folklore with dynamic social conditions.

Local cultures in regions of significant seismic

activity around the world are rich in myths,

legends, and other symbolic representations of

From: PICCARDI, L. & MASSE, W. B. (eds) Myth and Geology.

Geological Society, London, Special Publications, 273, 67–94.

0305-8719/07/$15.00 # The Geological Society of London 2007.


psychology vis-a`-vis the violent forces of nature,

and other aspects of society and culture. This lore

can also shed useful light on the geological

record, sometimes even to the extent of suggesting

major geological events that remain undiscovered

by conventional scientific approaches. Common

themes appear in stories from different cultures,

and may help identify stories with geological


In this paper, we examine two types of earth-
quake lore from Cascadia and Japan. First, we

discuss figurative stories from the Pacific Northwest

coast of North America that appear to refer to earth-
quakes, tsunamis, permanent land level changes, or

landslides. Geographically these stories describe

events along two major fault zones; the Cascadia

subduction zone (CSZ), which produced a magni-
tude 9þ earthquake in 1700 (Satake et al. 2003),

and the Seattle fault in Puget Sound which produced

an earthquake of estimated magnitude 7.4 in

approximately 900 AD (Bucknam et al. 1992). Sec-
ondly, we discuss non-geological evidence from

Cascadia and Japan that researchers have used to

date the CSZ earthquake of 1700. Next, we

examine figurative conceptions of earthquake caus-
ality in Japanese folk culture, both circa 1700 and,

in greater detail, during the period following the

Edo ( present-day Tokyo) earthquake of 1855.

This earthquake produced an outpouring of figura-
tive namazu-e (catfish picture prints), which

expressed a wide range of popular views on earth-
quake-related phenomena, both geological and

social. Data from both Cascadia and Japan

support our general argument that symbolic

language can usefully describe geological events.

In addition to demonstrating a linkage between

local earthquake lore and geological events in these

two parts of the world, we propose some observations

about similarities in this lore, with reference to other

regions of the world. At a deep level, which we call

the ‘archetypical level’, many apparently uncon-
nected societies throughout the premodern world

conceived of earthquakes in similar ways.

Stories of earthquakes and related events

from native societies in the Cascadia

subduction zone

Geological knowledge of the Cascadia

subduction zone

The plate-boundary fault at the Cascadia subduction

zone (CSZ) separates the oceanic Juan de Fuca plate

from the continental North America plate (Fig. 1). It

lies about 80 km offshore and extends roughly

parallel to the coast from the middle of Vancouver

Island to northern California. Although recognized

Fig. 1. Estimated 1700 rupture of the Cascadia

Subduction zone, from Wang et al. (2003). Numbers

indicate locales of Native stories with descriptions of

shaking and/or flooding. Story elements are tabulated in

Figure 9. Story references: 1. Boas 1935, 1a 33; 1b 92;

1c 122; 1d 27–31; 2. Teit 1912, 273–274; 3. Jenness

1955, 11,12,72,91,92; 4. Duff 1955, 9, 123–126; 5.

Roberts & Swadesh 1955, 315; 6. Sproat 1987, 124–

125; 7. Arima et al. 1991, 230–231; 8. Hill-Tout 1978,

157–158; 9. McCurdy 1961, 109–112. 10. Swan 1870,

57; 11. Gunther 1925, 119; 12. Clark 1953, 44–45; 13.

Eels 1878; 14. Reagan and Walters 1933, 14a 320–321,

14b 322; 15. Reagan 1934, 15a 33–34, 36–37; 16.

Jefferson 2001, 69–70; 17. Elmendorf 1961, 133–139;

18. Van Winkle Palmer 1925, 99–102; 19. Clark 1955,

321; 20. Boas 1894, 144–148; 21. Kuykendall 1889, 67;

22. 22a Boas 1898; 23–27 (similar story identified as

historical in the following reference), 22b 30–34, 22c

140; 23. Jacobs 1959, 176; 24. Jacobs 2003, 187; 25.

Frachtenberg 1920, 67–91; 26. Frachtenberg 1913, 14–

19; 27. Jacobs 1939, 58; 28. Ward 1986. 27; 29. Dubois

1932, 261; 30. Spott & Kroeber 1942, 224–227; 31.

Kroeber 1976, 31a 174–177; 31b 460–465; 32.

Warburton and Endert 1966, 58–60.


as early as the mid-1960s, seismologists initially

assumed that the CSZ was incapable of producing

great (megathrust) earthquakes. It is seismically

quiet, and no sizable earthquake has occurred on

it since European settlement began. As the theory

of plate tectonics matured, studies of subduction

zones worldwide identified characteristics associ-
ated with megathrust earthquakes. These earth-
quakes are most common in areas where hot,

young, buoyant crust is rapidly subducted (Heaton

& Kanamori 1984). Although the rate of subduction

in Cascadia is relatively slow, the subducted crust is

among the youngest and hottest anywhere.

Field investigations in the 1980s of the coastal

margins along the CSZ located geological evidence

of abrupt land-level changes characteristic of mega-
thrust earthquakes in ‘ghost forests’ of dead cedar

trees in coastal estuaries in Washington and

Oregon (e.g. Nelson et al. 1995). The cedars, orig-
inally above the limit of the tides, were killed when

their roots were suddenly plunged into salt water.

Beneath the surface of these same estuaries, soil

cores reveal layered deposits showing a repeated

cycle of slow uplift and rapid submergence. Pre-
liminary age estimates based on radiocarbon

dating (Nelson et al. 1995) and tree-ring studies

suggested that the most recent earthquake happened

about 300 years ago. A precise date, 26 January,

1700, was determined from Japanese historical

documents (Satake et al. 2003), and confirmed by

a close study of tree-ring patterns of ghost cedar

roots (Yamaguchi et al. 1997). The magnitude esti-
mate of 9.0, derived from the amplitude of the

tsunami that reached Japan, implies rupture along

the entire length of the CSZ (Satake et al. 2003).

Figure 1 shows the geographic extent of the likely

rupture area.

Native folklore from the Cascadia

subduction zone

This section examines Native stories from along the

Cascadia margin that are figurative and folkloric in

style, and not amenable to dating with any pre-
cision. Some of these stories appear to be of fairly

recent origin and possibly linked to the 1700 earth-
quake; others are apparently much older.

Native peoples have lived on the Cascadia coast

for thousands of years, transferring knowledge

from generation to generation through storytelling.

These Native groups spoke more than a dozen dis-
tinct languages (Thompson & Kinkade 1990), and

lived in a complex social landscape with both simi-
larities and differences between groups. Collection

and recording of Native stories began in the

1860s, almost 100 years after initial European

contact in Cascadia, resulting in losses of Native


Thunderbird and Whale are beings of superna-
tural size and power. A story from Vancouver

Island says that Thunderbird causes thunder by

moving even a feather, and that he carries a large

lake on his back from which water pours during

thunderstorms (Carmichael 1922). The same story

says that all creation rests on the back of a

mammoth whale and tells of an occasion when

Thunderbird drove his talons deep into the quivering

flesh of Whale’s back, and Whale dived and dragged

the struggling Thunderbird to the bottom of the

ocean; imagery suggestive of ground shaking and

ocean surges. In this story, three of the four original

thunderbirds were drowned in this manner, and one

remains alive. Other stories also have multiple

whales or thunderbirds (Fig. 1, stories 1d, 15b,

22b; Reagan 1934, p. 25; Spott & Kroeber 1942,

p. 227–232) that may refer to aftershocks.

Stories 5, 9, 14a and b, and 15a (see Fig. 1) further

tie the story of a supernatural battle to the flood, with

imagery that implies shaking—Thunderbird lifts the

massive Whale into the air and drops it on the land

surface. The flood description in story 15a is strik-
ingly similar to story 10, which hints at a historic

framwork by placing the event ‘A long time ago

… but not at a very remote period’.

The struggle between Thunderbird and Whale is

unique to the Cascadia coast, and appears in stories

from Vancouver Island to northern Oregon. From

central Oregon south, Thunder or Whale figures

appear individually in stories describing earthquake

or tsunami themes. The central figures variously

appear in the form of Thunder, Thunderbird or

bird, and Whale, fish, or sea monster. In northern

California, the Yurok tribe has an ‘Earthquake’

figure with ‘Thunder’ as his companion. Stories

from Puget Sound and eastern Washington also use

similar motifs in conjunction with descriptions of

earthquake effects. Story 16, of the battle between

the double-headed eagle and the water-monster, is

about the creation of Agate Pass, a Puget Sound

waterway far from the outer coast, but adjacent to

the Seattle Fault, where a magnitude 7.4 earthquake

caused a Puget Sound tsunami (Moore & Mohrig

1994) about 1100 years ago (Bucknam et al. 1992).

are dateable, a few have vaguely historical time-
frames. In addition to describing earthquake

effects, Thunderbird and/or Whale stories have a

general association with landscape-forming

events, such as glacial moraines (Fig. 1, story

15b), icefalls (Reagan & Walters 1933), and land-
slides (Barbeau & Melvin 1943). Thunderbird also

appears in stories about thunder, lightning, and

rain. Thunderbird and Whale stories are part of a

systematic oral tradition that used symbolism and

mnemonic keys to condense and present infor-
mation in a format that could be remembered and

retold for generations. Native populations wit-
nessed multiple cycles of CSZ earthquakes; geo-
logical evidence indicates at least seven in the last

3500 years (Atwater & Hemphill-Haley 1997).

Artifacts depicting Thunderbird and Whale that

long predate the 1700 earthquake have been recov-
ered from coastal archeological sites (McMillan

2000). Knowledge of a repeating earthquake cycle

may be implied in a story where the Thunderbird

becomes a man and sends his Thunderbird

costume back to the sky saying: ‘You will not

keep on thundering, only sometimes you will

sound when my later generations will go (die).

You will speak once at a time when those who

will change places with me will go (die)’ (Boas

1935, p. 65).

theme in carved and painted art of the outer coast

and coastal fjords of Vancouver Island (Malin

1999) (Figs 2 & 3), where broad ocean openings

Fig. 2. Two interior ceremonial screens from Port Alberni, dating from the late nineteenth century. The screens

depict the Thunderbird, accompanied by the lighting serpent and wolf, carrying the Whale in its talons. Collection of

American Museum of Natural History; 16.1/1892 AB. The screens are said to commemorate a ‘chief’s encounter

with the supernatural while checking his sockeye traps at Sproat Falls’ (Kirk 1986). Sproat Falls is just above

the modelled extent of the 1700 tsunami (Clague et al. 2000).


Fig. 3. Nootka Sound Memorial, erected 1902–1903 to

honour Chief Maquinna, who died in 1902. Thunderbird

and Whale are shown as similar in size to Conuma Peak.

Photo PN11478-A, taken by C.H. French and

reproduced with the permission of the Royal British

Columbia Museum.

funnel water into narrow waterways that run far

inland. Port Alberni, at the landward terminus of

Barkley Sound, 40 km from the ocean, experienced

tsunami run-up about six times larger than sites on

the open coast following the 1964 Alaska earth-
quake (Sokolowski Alaska Tsunami Warning

Centre). Clague et al. (2000) have documented

tsunami deposits from both the 1964 and 1700

earthquakes in Port Alberni and other fjord-like

inlets on Vancouver Island. Alert Bay, between

the northern end of Vancouver Island and the main-
land, also has prominent Thunderbird and Whale

artworks (Fig. 4) and story themes linking Thunder-
bird and flooding (Fig. 1, story 1a), and placing

flooding at the time of the winter ceremonial

(Fig. 1, story 1b).

Native stories, art, ceremonies, and naming pre-
serve memories of Cascadia subduction zone earth-
quakes. Ancient, recurring imagery describes

earthquake and tsunami effects and suggests aware-
ness of repetitive cycles of world-altering events.

Likewise, similarities in symbols and imagery

along the length of Cascadia suggest commonly

experienced events. We now take a closer look at

earthquake-related lore from the Puget Sound area.

A’yahos, the AD 900 Seattle earthquake and

earthquake lore from the Puget Sound area

along the Seattle fault

The Seattle fault is a multi-stranded east –west

striking reverse fault cutting across Puget Sound,

through downtown Seattle, and across Lake

Washington. Although geophysical evidence has


Blukis Onat 1987). Natives passed down knowl-
edge of these events in their oral tradition using

descriptive metaphors based on their cultural con-
cepts, often ascribing earth shaking to actions of

supernatural beings.

In 1985, prior to published evidence of the AD

900 earthquake on the Seattle fault, an article in

the Seattle Weekly (Buerge 1985) mentioned a

Native American ‘spirit boulder’ associated with

earthquakes and landslides located near the Fauntle-
roy ferry dock in west Seattle. The proximity of this

location to the Seattle fault invited investigation and

we discovered that the Fauntleroy Spirit boulder is

associated with a supernatural being called

a’yahos. Native stories often describe a’yahos in a

way that could refer to earthquake effects,

especially landslides. A’yahos is a shape-shifter,

often appearing as an enormous serpent, sometimes

double headed with blazing eyes and horns, or as a

composite monster having the fore-quarters and

head of a deer and the tail of a snake (Mohling

1957). A’yahos is a ‘Doctor’ spirit power; reserved

for shamans. It is one of the most powerful personal

spirit powers; malevolent and dangerous to encoun-
ter. A’yahos is associated with shaking and rushes

of turbid water and comes simultaneously from

land and sea (Smith unpublished notes). ‘At the

spot where a’yahos came to a person the very

earth was torn, landslides occurred and the trees

became twisted and warped. Such spots were recog-
nizable for years afterward.’ (Smith 1940)

specific places in the central Puget Sound, along

the Hood Canal, and on the Strait of Juan de Fuca

as far west as the Elwha River. A total of 13

a’yahos sites are mentioned in various stories

(Fig. 5a, b), and these locales coincide with

shallow faults around Puget Sound, including the

Little River fault along the strait of Juan de Fuca,

the Tacoma fault, and the Price Lake scarps

(Haugerud et al. 2003). Five of the a’yahos story

sites are located very close to the trace of the

Seattle Fault (Fig. 5b). Four of the Seattle locales

can be associated with landslides or reports of

land-level changes that might have been caused

by the AD 900 Seattle earthquake. Additional Native

stories related to shaking, landsliding, or land-level

change are associated with three of these sites.

A’yahos stories along the Seattle fault

The west Seattle a’yahos spirit boulder mentioned

by Buerge (1985) is located on the beach immedi-
ately south of Fauntleroy Ferry Dock (Fig. 5b:1),

Fig. 5. (a) Map of Puget Sound and eastern Olympic Peninsula. Boxed area indicates location of larger-scale map

shown in Figure 5b. Dashed lines show locations of some shallow faults (after Haugerud et al. 2003); LR F, Little River

fault; T F, Tacoma Fault; DDM FZ Darrington Devil’s Mtn fault zone; PL S, Price Lake Scarps; FC S Frigid Creek Scarps.

Numbers in Figure 5a indicate sites outside the Seattle fault area associated with a’yahous stories. 1, Elwha River; 2,

Dungeness River; 3, Dabob Bay; 4, Bald Point also known as Ayers Point; 5, Tahuya River; 6, Medicine Creek (Nisqually

Delta); 7, American Lake; 8, Black Diamond Lake (1–5 from Elmendorf, 1993; 6 and 8 from Waterman 2001; 7 from

Smith, 1940). (b) Larger-scale map showing the Seattle fault zone, a’yahos story localities (indicated by black circles),

other stories that have apparent connection to earth shaking or landsliding (indicated by grey circles), and archaeological

sites (white circles). 1, Fauntleroy; 2, Alki Point; 3, Lake Washington a’yahos site; 4, South Point, Mercer Island; 5,

Madison Park; 6, Three Tree Point; 7, Agate Passage; 8, Bremerton; 9, Moore Point; 10, Portage Bay; 11, West Point; 12,

Duwamish Site No. 1. LIDAR images of Fauntleroy (1) and Three Tree Point (6) are shown in Figure 6.


below what appears to be a very large landslide of

undetermined age clearly visible in LIDAR

images (Fig. 6a) but not shown on existing geologi-
cal maps. Long term local residents Mory Skaret

and Judy Pickens pointed out the boulder; Water-
man (2001) indicated a location further south,

near Brace Point. Stories of a’yahos spirit power

are told about both the Fauntleroy boulder (Water-
man 2001) and Alki Point (Smith unpublished

notes), immediately to the north and uplifted

during the AD 900 quake. Stories about Alki Point

speak of shaking, rocks exploding, and the power

coming from sea and land simultaneously (Smith

unpublished notes).

The second place in Seattle associated with

a’yahos is by the shore of Lake Washington (Fig.

5b: 3). According to elders who worked with T.T.

Waterman, ‘On the lake shore opposite the north

end of Mercer Island … an enormous supernatural

monster … lived’ (Waterman 2001, p. 102). Large

block landslides dated to AD 900 slid into Lake

Washington from the southern end of Mercer

Island and at Madison Park (Karlin & Abella

1992), about 2 km south and north, respectively,

Fig. 6. LIDAR images (from the Puget Sound LIDAR

Consortium 2000) showing apparent landslides at

localities said to be a’yahos dwelling places; (a) Fauntleroy

Cove in West Seattle (b) Three tree point in Burien.


The description of the widened channel could

reflect permanent ground level change, and the

sense of ground motion suggested by the story is

accurate; Agate Passage is on the down-thrown

northern side of the Seattle fault. However, geologi-
cal evidence suggests that the AD 900 earthquake

produced mainly uplift on the southern side, with

the north side down only slightly; the correspon-
dence between the story and reality is approximate

rather than exact. We note, however, that some

‘drift’ seems reasonable in a story that may be a

thousand years old and has been preserved through

extreme cultural destruction. This story, set in an

undated ‘long ago’, is strikingly similar to the

stories from the outer coast of Cascadia that use

the struggle of a supernatural bird and water-beast

to refer to earthquakes on the Cascadia subduction

zone (Ludwin et al. 2005a). The ‘long ago’ time

frame suggests an origin more ancient than 1700.

A fifth place, on the Kitsap Peninsula near

Bremerton (Fig. 5b: 8), is said to be another spot

where shamanistic spirit-power could be acquired

(Waterman 2001, pp. 206 – 207; Smith unpublished

notes). Sam Wilson, born in 1861, and grandson of

Chief Seattle told Marian Smith, ‘it comes from

land and sea at same time’ (Smith unpublished

notes). No obvious geological features were noted

at this site, though it is situated between several

strands of the Seattle fault. On the Puget Sound

shore of Kitsap Peninsula just east of this locality,

at Moore Point near Illahee State Park (Fig. 5b: 9),

is a spot named ‘to have a chill’ or ‘to feel a

tremor’ (Waterman 2001, pp. 206 – 207). A com-
parison of earth tremors to feverish chills was

made by Aristotle (Leet 1948) and it is possible

that the Natives of Puget Sound drew a similar con-
nection. Although the origin of the name ‘to feel a

tremor’ is uncertain, shaking was a central

element in Puget Sound Native medical practices

and ceremonials, and a’yahos was a potent source

of shamanistic ‘Doctor’ power, as discussed below.

Native ‘Doctor’ or shaman power was a particu-
larly strong form of spirit power. Throughout the

region, individuals sought personal spirit powers to

guide their lives and bring them luck and skill.

A’yahos was one of the most powerful of these per-
sonal spirit powers, though it was also malevolent,

dangerous, and possibly fatal to encounter (Smith

1940). A’yahos ‘Doctor’ spirit power was one of

only two powers (a’yahos and sta ́dukw’a) reserved

exclusively for shaman, and descriptions of both

these shamanistic powers include shaking or land-
sliding imagery (Elmendorf 1993; Smith unpub-
lished notes; Smith 1940; Waterman 2001).

Shaman were believed to hold the power to cure

certain illnesses, and also the power to cause illness

and even death (Suttles & Lane 1990). The name

of James Zackuse, a Duwamish Indian Doctor who

lived in Seattle on Lake Union’s Portage Bay

during the late nineteenth and early twentieth centu-
ries, translates to ‘trembling face’; rooted in ‘dzakw’,

the Puget Lowland Native word for earthquake

(Miller & Blukas Onat 2004, pp. 78–85).

Puget Sound Salish ceremony, when ritual objects

filled with spirit power and became self-animated

(Miller 1999, p. 133; Elmendorf 1993, p. 192 – 198;

Haeberlin & Gunther 1930, p. 79). An early white

settler noted a specific connection between cer-
emony and earthquake shaking as early as 1893:

During the past thirty-three years I have on many occasions endea-
vored to gather from the oldest and most intelligent Indians some-
thing of their earlier recollections; for instance, as to when the

heaviest earthquake occurred. They said that one was said to

have occurred a great many years before any white man had

ever been seen here, when mam-ook ta-mah-na-wis was carried

on by hundreds. This is the same performance they go through

when they are making medicine men, and consists of shouting,

singing, beating on drums and sticks and apparently trying to

make as much noise as they can. (Seattle Post-Intelligencer 1893)

Salish earthquake stories from outside Puget Sound

also draw a connection between ceremony and

shaking (Fig. 1, stories 8 and 22b; Ludwin et al.


Earthquake lore from Puget Sound in the

context of regional earthquake motifs

Although the a’yahos name appears to be specific to

central Puget Sound, the double-headed serpent is

widely known and depicted in NW cultures, and

may have been similarly linked to earth changes.

The Quileute, a non-Salish group living on the NW

Washington Coast, have artifacts depicting a two-
headed horned snake with the forelegs of a deer.

Although not clearly linked to a’yahos, stories

describe it as a vicious guardian spirit (Powell &

Jensen 1976). Another two-headed snake, the

Sisiutl, is a figure well known from stories and

ceremonial artifacts of northern Vancouver Island.

the subterranean world in the same way that snakes

do) appear in many Pacific Northwest coastal

stories that describe ground shaking and/or

tsunami-like floods, probably related to earthquakes

on the Cascadia subduction zone (Ludwin et al.

2005a). Whales per se are not prominent in

stories from the Seattle fault area, though the

water-serpent of Agate Pass is analogous to a

whale. However, in southern Puget Sound where

damaging earthquakes centred deep underground

are relatively common (occurring in 1949, 1965,

and 2001), several stories mention whales trapped

inland and thrashing their way out, sometimes

through underground channels (Ballard 1929).

Thunder, also common in coastal stories of


shaking and flooding, appears occasionally in

stories from Puget Sound (Ludwin et al. 2005b).

Figure 7 shows two versions of a Salish ceremo-
nial dance mask and costume linked to earthquakes

(Le ́vi-Strauss 1979), whirlwind (American

Museum of Natural History catalog), and thunder

(Jenness 1955). The Sxwayxwey (also Swai’xwe

and many alternate spellings) masks sometimes

include a two-headed snake (Jenness 1955). The

mask’s origin is relatively recent, probably some-
time after 1500 (Ludwin et al. 2005a), and is

described in a number of Salish stories that use

Fig. 7. Salish Swai’xwe masks associated with shaking,

whirlwind, thunder and the two-headed snake (Jenness

1955). The two open-mouthed protuberances above the

forehead likely represent snakes. (a) Mask from

mainland British Columbia, collection of American

Museum of Natural History; 16/9222A. (b) Mask from

Vancouver Island, photo by Edward Curtis (2001).


Fig. 8. Non-Salish Cascadia Native representations of

two-headed snakes. Neither of these figures has yet been

explicitly linked to earthquakes, but they likely represent

the same spirit power as a’yahous. Both have horns,

representing spirit power. (a) Quileute ceremonial

representation of t’abale, a vicious guardian spirit on the

northwestern Washington coast (Powell & Jensen 1976).

(b) Kwakwaka’wakw Sisiutl mask, from the northern end

of Vancouver Island, photo by Edward Curtis (2001).

Cannibal the additional names Rolling-Down,

Great-Mountain, Rock-Slide and Coming Down.

The two-headed Sisutl of the Kwakwaka’wakw is

similar in form to the two-headed supernatural

serpent a’yahos of Puget Sound, and its blood trans-
forms the child of the Thunderbird/Dzonoqwa into

the earthquake-related figure Stone-Body. The

inclusion of multiple earthquake-related mythic

figures (Thunderbird, Dzonoqwa, Stone-Body,

Sxwayxwey, Sisiutl) in a story about the foundation

of the great houses of the Kwakwaka’wakw

suggests that earthquakes deeply affected their

culture. The use of earthquake imagery from the

adjoining Salish and Haida cultures suggests earth-
quake events that were felt across tribal boundaries.

Non-geological evidence for the Cascadia

subduction zone earthquake of 1700 from

Cascadia and Japan

The precise dating of the Cascadia subduction zone

earthquake of 1700 is an example of how local lore

and other non-geological evidence can enhance

conventional geological knowledge. The 1700

earthquake was initially dated through Japanese his-
torical documents, and the date was confirmed inde-
pendently through Native American oral traditions

and dendrochronology.

(Fig. 1, stories 1c, 3, 4, 6, 7, 13, 17, 27 and 28)

have sufficient information for estimating a date

range since an event associated with shaking and/

or flooding (two stories with both, three with

shaking only, and four with flooding only). Two

stories, told between 1860 and 1964, tell of a grand-
parent who saw a survivor of the flood, and one of a

great-grandparent who survived it. Figure 9 tabu-
lates the accounts, and gives date ranges. Date

range minima and maxima are 1400 and 1825. All

estimates span the interval between 1690 and

1715, and the average value of the midpoints of

the date ranges is 1690. Discarding the earliest

and latest midpoints yields an average midpoint

date of 1701. This finding is remarkably consistent

with the 1700 date of the most recent CSZ earth-
quake determined from Japanese historical


of floods could possibly be reports of tele-tsunamis

(i.e. those arriving from distant earthquakes).

Alaskan and South American earthquakes produced

notable tsunamis on the Cascadia coast in the

twentieth century (Lander et al. 1993). Although

we do not know the history of Alaskan earthquakes

around 1700, tsunamis from South American earth-
quakes were recorded in Japan in 1730, 1751 and

1780 (Watanabe 1998). Japanese earthquakes

have not produced significant tsunamis in Cascadia

since at least 1806 (Lander et al. 1993), but locally

generated tsunamis damaged the Japanese coast in

1611, 1707, and 1771 (Watanabe 1998).

the 1700 earthquake are mostly straightforward

descriptions of flooding and/or shaking. Of these

stories, the clearest and most complete is from the

outer coast of Vancouver Island, recorded by

Chief Louis Nookmis following the 1964 Alaskan

earthquake. It describes a night-time earthquake

quickly followed by a tsunami that destroyed the

Pachena Bay people:

They had practically no way or time to try to save themselves. I

think it was at nighttime that the land shook … . I think a big

wave smashed into the beach. The Pachena Bay people were

lost … . But they who lived at Ma:lts’a:s, ‘House-Up-
Against-Hill’ the wave did not reach because they were on high

ground … Because of that they came out alive. They did not

drift out to sea with the others. (Fig. 1, story 7, Arima et al. 1991)

Huu-ay-aht First Nation and descendent of Chief


Fig. 9. (a) Tabulation of story elements for stories listed in Figure 1; effects, figurative motifs, and environmental

setting. Brackets by story numbers group stories from a common geographic locale. Symbols are as in Figure 1.

The ‘Whale’ motif is enclosed in quotes to cover a variety of sea-monsters appearing in the stories. (b) Date

range estimates and quotes used to calculate date range estimates. Date range estimates used the following

assumptions: a ‘generation’ is no fewer than 15 and no more than 40 years, events before age 5 are not remembered,

the maximum lifespan is 100 years, flood survivors were ‘old’ when seen, and an ‘old’ person is at least 40.

Louis Nookmis, has discovered previously unpub-
lished information that allows us to estimate a

date at between 1640 and 1740. This new infor-
mation comes from a comprehensive transcription

and translation of the 1964 recordings undertaken

by the Huu-ay-aht First Nation.

A second datable story that includes flooding and

shaking elements is from the northern margin of the

Olympic Peninsula in northwestern Washington. It

combines information from three independent

sources (Fig. 1, stories 11– 13) to yield a tale indi-
cating winter flooding with accompanying strong

shaking. A tradition that cannot be dated but

vividly describes strong night-time shaking, from


abandoned following the 1700 earthquake and

tsunami (Minor & Grant 1996; Hutchinson &

McMillan 1997; Losey 2002; Cole et al. 1996),

supporting the possibility that flooding stories

may reflect this event.

As we mentioned earlier, Japanese textual data

were instrumental in precisely dating the CSZ earth-
quake of 1700. The exact date and approximate time

of this earthquake (9 pm on 26 January 1700) were

determined from a variety of Japanese historical

documents such as domain (han) records, merchant

records, and the records of village headmen that

reported the arrival of a tsunami with no reports of

associated shaking (Satake et al. 2003). In addition

to recording the 1700 earthquake, Japan has a rich

folklore related to earthquakes and written and

graphic documentation that allows us to observe

how that folklore developed and interacted with

other aspects of Japanese culture. Earthquake

imagery in Japanese folklore has distinct similarities

to Cascadia imagery, and we explore this, particu-
larly through the example of 1855 Ansei earthquake,

which was followed for a few months by a brief but

abundant output of ‘namazu-e’ (catfish picture-
prints) that combined earlier earthquake folklore

with incisive observations on both earthquake

effects and current events.

Halfway to the present and halfway around

the world—The 1855 Ansei earthquake

in Japanese folk images

Japanese documents used to date the 1700 earthquake

focus on straightforward descriptions of areas flooded

by the 1700 tsunami and resultant damage and do not

touch upon the origin of the event. However, Japan

lies in an area of especially vigorous seismic activity

and it is not surprising that we can find abundant

earthquake-related data expressed both as written

records describing the effects of specific events and

in folk culture ideas about their cause. The long

written history available in Japan enables us to

track changing conceptions of earthquakes and

offers an interesting comparison to the earthquake

stories from the oral traditions of Cascadia. For

example dragons and other serpent-like creatures

associated with water were prominent in Chinese

and Japanese folk beliefs concerning earthquakes.

Figure 10 shows a broadsheet entitiled ‘The cause

of earthquakes and tsunamis’ published c. 1650. In

Japan, the serpent figure gradually gave way to that

of a giant catfish (namazu), a belief that parallels

the many shaking-related whale stories found in the

Pacific Northwest (Ludwin et al. 2005a).

The link between earthquakes and giant catfish

developed gradually over several centuries from

native Japanese folk beliefs with some influence

Fig. 10. ‘Earthquakes and Tsunamis Explained’,

c. mid-seventeeth century. On the outer edges of the

circled dragon are written the months of the year. What

appears to be a small sword is just above and touching

the dragon’s head. Below this sword is written

‘kaname-ishi’, (foundation stone). Inside the dragon are

the ‘the 60 plus islands of Japan and the various foreign

countries’. The last line of text inside the dragon

explains that all of these places should be regarded as

existing above the dragon. In other words, the dragon

resides under the earth. Normally, it is pinned down and

made immobile by the deity of the Kashima Shrine, who

presses down on a boulder (the foundation stone), which

presses down on the dragon’s head. The deity’s sword is

a substitute for the boulder. Sometimes, however, the

deity dozes or is otherwise distracted, and he lets up on

the boulder. The dragon is thus able to wiggle around

under the earth, which causes earthquakes (from Miyata

& Takada 1995, p. 54).

of Chinese ideas. The basic view was that a giant

namazu lived in the subterranean waters below the

Kashima Shrine in Hitachi Province (present-day

Ibaraki Prefecture, slightly north of Tokyo). A

large boulder called the foundation stone

(kaname-ishi) pinned the namazu down and kept

it largely immobile. The weight of the foundation

stone itself, however, was insufficient to suppress

the namazu’s movements, and the system depended

on the Kashima deity (Kashima daimyo ̄jin, often

known simply as Kashima) pressing down on the

stone. During the tenth month of each year

Kashima had to leave his post and travel south to

Ise to attend a meeting of the major Japanese

deities. In his absence, Kashima would leave the

local deity Ebisu in charge of pressing down on

the foundation stone. Whether owing to negligence

by Kashima himself or to Ebisu’s inability to

perform the namazu suppression tasks, earthquakes

took place when the lack of pressure on the

foundation stone allowed the giant namazu to

wiggle around under the earth. The severity of

shaking depended on the extent of the namazu’s



This basic understanding of the namazu-based

cause of earthquakes was subject to many variations

because it was enmeshed in the broader network of

Japanese folk religion. Cornelis Ouwehand’s

detailed, structuralist study of namazu images situ-
ates their themes within the broader matrix of folk

religion (Ouwehand 1964). One twist on the basic

motif was that Kashima often worked in close associ-
ation with the thunder deity and sometimes other

local deites of Edo. Namazu-e sometimes depicted

Kashima, Ebisu, and the thunder deity as being

jointly responsible for the devastation of a major

earthquake. Also, most early nineteenth-century

Japanese people associated earthquakes with water.

The namazu, of course, was a water-dwelling crea-
ture and the thunder deity manifests himself in

storms. Indeed, most popular newspaper accounts

of earthquakes also mention the presence of thunder-
storms associated with them (e.g. Kitahara 1999,

pp. 32–33, 36–37).

Although the namazu-based explanation of earth-
quakes had become widely known throughout

Japan by the early nineteenth century, it was not

the only way of describing the mechanism of earth-
quakes. The Ansei kenmonroku (Accounts of the

Ansei [1854 – 1859] era) contains a typical alterna-
tive, based on a widely known view of cosmic trans-
formation whereby the five primary agents of yin

and yang—fire, metal, wood, earth, water—

interacted to create the material world and to

embody the forces that govern it. With respect to

earthquakes, normally water ( purely yin) over-
comes fire (purely yang). Furthermore, water is

the agent normally holding sway in the subterra-
nean environment. Earthquakes result from the

occasions when fire overcomes water underground,

thus reversing the normal state of affairs. A broad-
sheet issued just after the Ansei earthquake of

1855 explained its cause in terms of both yin and

yang forces and the movements of namazu, but it

called the namazu-based explanation an ‘unsophis-
ticated theory’. (Wakamizu 2003, pp. 16 – 17).

Popular newspapers often started their accounts of

earthquakes with a simple, brief statement of yin

and yang forces being out of balance. For

example, the text of an account of the Ise earth-
quake (14th day, 6th month, 1854) explains that a

clash of yin and yang forces resulted in thunder in

the skies and shaking of the earth. An account of

an earthquake in Odawara (2nd day, 2nd month,

1853) employs verbatim the same explanation

(Kitahara 1999, pp. 32 – 33).

The key point here is that in nineteenth-century

Japan, multiple theories of earthquake causality

co-existed. Most of these theories postulated an

imbalance in the cosmic forces, expressed in

terms of the five agents (gogyo ̄) of yin and yang

or the subterranean movement of a giant creature.


Fig. 11. Untitled namazu-e showing (1) the

co-existence of two modes of thinking regarding the

causes of earthquakes and (2) the namazu as an agent of

world rectification ( yo-naoshi). Three members of the

construction trades, identified by their tools, are

celebrating their newfound wealth (the gold coins

apparently falling from the sky) by drinking with the

namazu. The foundation stone appears to be floating in

the air. On jacket of the man in the left foreground is the

character for earth ( ), while the jacket of the man in

the right foreground reveals the character for fire ( ).

The character for water ( ) forms the pattern of the

namazu’s robes, and the character for wood ( ) does

the same for the jacket of the man behind the namazu.

The airborne gold coins stand for metal ( ), whose

character also means gold or money. Earth, fire, water,

wood, and metal are the five agents of yin and yang,

whose imbalance was the cause of earthquakes in many

premodern theories throughout East Asia. The shaking

of the earthquake rectifies this imbalance, both in an

abstract sense and in more specific ways. In this case, the

tradesmen are receiving metal (gold, money) from the

wealthy members of society. Here the namazu can be

viewed as a literal cause of earthquakes, as a metaphor

for earthquakes, and as a symbol of social rectification

(from Wakamizu 2003, p. 69).

in the forces of yin and yang. Nevertheless, the

close link between namazu (or anything similar)

and earthquakes never developed in China.

Perhaps the most significant Chinese influence on

Japanese views of earthquakes came from the

ancient idea of heaven’s mandate (tianming). In

this view, which could accommodate both abstract

and anthropomorphic conceptions of the cosmic

forces, heaven (the cosmos) bestows on rulers a

mandate to govern based on their moral fitness.

Earthquakes, floods, famine, epidemics, and other

natural calamities were signs of heaven’s displea-
sure. This idea became the bedrock of classical

Chinese political theory. It was also influential in

Japan, especially in the notion that the cosmic

forces periodically rectify a social order gone

awry ( yonaoshi, ‘world rectification’). Earthquakes

were a major tool for bringing about such rectifica-
tion, and in this sense, they were not random occur-
rences. The print described above in which the

earthquake redistributes wealth reflects this way

of thinking. Earthquakes, therefore, necessarily

had political significance in premodern Japan, and

commentary on them could easily become com-
mentary on the state of society and government.

The namazu-e (catfish picture prints):

Japanese responses to the Ansei earthquake

For Japan, a particularly well documented example

of how folk beliefs intersected with contemporary

political and social culture is the Ansei earthquake

of 1855. On the second day of the tenth month

(November 11 in the solar calendar), a magnitude

6.9 earthquake with a shallow focus shook Edo

(present-day Tokyo) and a wide surrounding area.

Aftershocks continued for the next nine days.

Estimates of the number killed in the greater Edo

area range from 7000 to 10 000 (4000 – 5000 for

the downtown area), but the precise figure is uncer-
tain. This death toll amounted to roughly 1 in 170

Edo residents, and shaking and subsequent fires

destroyed 1 in 3 non-military houses and other

structures (Inagaki 1995, p. 64). The injured were

especially numerous, and fires burned for days

throughout the city.

geography, and politics magnified the psychologi-
cal impact of this earthquake in such a way as to

make it appear as a direct attack on the heart of

the bakufu, Japan’s military government based in

Edo. The distribution and severity of damage was

not uniform. Some areas suffered severe devas-
tation and loss of life, whereas other parts of the

city came through the ordeal with nearly all build-
ings and people shaken but intact. The damage

was less a function of proximity to the epicentre

than it was a function of topography and soil con-
ditions. The Yamanote Tablelands, an extension

of the Musashino Plateau, wound their way

through parts of the heart of Edo, constituting

modest upland areas. These upland areas were not


always obvious because of erosion and past filling

with soil or debris of low-lying areas. In 1590,

when Tokugawa Ieyasu (1542– 1616) made the

fishing village of Edo his base of operations,

human engineers and construction workers began

to reclaim the marshy flats around Edo Castle.

This process accelerated rapidly during the early

seventeenth century, after Edo became the de

facto political capital of Japan. Edo Castle itself

was on natural high ground, but much of the

prime land around the castle had been part of a

river drainage basin of Edo Bay a mere two or

three centuries earlier.

When the earthquake struck, it shook the whole

city, but structures on the firm foundation of the

uplands generally fared better. The severe damage

occurred in low-lying areas, especially areas of

land reclaimed from marshes and waterways. As

fate would have it, the most prominent neighbour-
hood of samurai residences, home to the bakufu’s

closest supporters among the domain lords,

leading bakufu officials, and several key bakufu

offices, was located at a place that during the

sixteenth century had been the Hibiya Inlet of Edo

Bay. The earthquake devastated this neighbour-
hood, as if it had targeted the government for

destruction. One residential zone further out from

the castle, the area adjacent to the elite neighbour-
hood, was home to commoners. Built on a firm

foundation, it suffered only moderate damage and

stood in stark contrast to the elite neighbourhood’s

collapse. In the eyes of commoners and elite alike,

the cosmic forces made a strong statement that

night (Noguchi 1997, pp. 73 – 108).

As if to add insult to injury, there was one more

odd twist to the earthquake damage. In the com-
moner neighbourhood of Kitachi-ku, for example,

not one main building collapsed. Nearly all the

serious injuries from this neighbourhood were the

result of falling roof tiles or eaves from collapsed

storehouses, built as separate structures from the

main buildings. Many other neighbourhoods reported

the same pattern, and all visual evidence points to

storehouses sustaining much worse damage than

any other type of structure. These rigid, heavy, mud

walled, tile-roofed storehouses tended to vibrate at

the same frequency as the high-frequency seismic

waves generated by the shallow-focus earthquake.

The irony is that the bakufu ordered this rigid,

heavy storehouse design in 1842 as a fire-prevention

measure (Noguchi 1997, pp. 118–120). In this way

too, the earthquake seemed to be paying especially

close attention to the government in its destruction.

Within two days of the initial shaking, printers

set up makeshift facilities in the relatively less

damaged areas and began to produce namazu-e

for sale through street vendors. Namazu-e sold

briskly for approximately two months before

Local folklore corresponds closely to geological evidence and geological events in at
least some cases, and the symbolic language of myth and folklore can be a useful
supplement to conventional geological evidence for constructing an accurate historical
record of geological activity. At a deep, archetypical level, Japan, Cascadia, and many
of the world’s cultures appear to share similar themes in their conception of
earthquakes. Although folklore from Cascadia is fragmentary, and the written record
short, the evolution of Japanese earthquake folklore has been well documented over a
long period of history and illustrates the interaction of folklore with dynamic social