Indigenous oral traditions of the Líl̓wat Nation recount observations of Qw̓elqw̓elústen (Mount Meager), a Garibaldi Volcanic Belt volcano in southwestern British Columbia, Canada; and associated eruptive activity, mass-wasting, and outburst flooding. We present Líl̓wat observations relating to Qw̓elqw̓elústen’s ∼2360 cal year B.P. eruption and its aftermath, a devastating outburst flood down the Lillooet valley. The Copper Canoe story correlates with the event sequence of pyroclastic damming of the Lillooet River and an outburst flood traveling far downstream, interrupting salmon runs and displacing people. Other stories suggest an eruptive plume and fumaroles. Recounted valley-floor changes, with proximal scouring and downstream filling of marshes allowing human resettlement, closely parallel and augment geological evidence, showing that oral traditions are equally important in holding landscape history. Oral traditions portray dramatic landscape changes, some by the Transformers, said to have traveled this land to make imperfect things right. Geologically documented debris-flow delta progradation and infill of the upper 50 km of Lillooet Lake since ∼12 000 cal B.P. underscore the land’s dynamism and the need for both sources to inform planning for future eruptive, mass-wasting, and flooding events. Traditional landscape knowledge, like Western science, is observational and evidence-based, though interpretations can differ given Indigenous belief in a sentient landscape, capable of acting with intention. Binding of stories to geographical locations has functioned as a powerful mnemonic device to preserve orally transmitted information across many generations.

The most recent volcanic eruption in southwestern British Columbia, Canada, occurred about 2400 years ago in the Mount Meager Volcanic Complex (MMVC), near the north extreme of the Garibaldi Volcanic Belt (GVB). Qw̓elqw̓elústen, as the MMVC is known to the Indigenous Líl̓wat Nation (Líl̓wat7úl), is a castellated massif signaled in oral traditions as the most powerful place in their traditional territory (Kennedy and Bouchard 2010) (Fig. 1A). We present evidence from oral traditions that Líl̓wat people witnessed Qw̓elqw̓elústen's eruption and experienced its devastating outburst-flood aftermath in the Lillooet River valley, and consider their accounts alongside the geological reconstruction. Our paper is structured in a stepped, alternating fashion between Indigenous and geological accounts, as a form of dialogue. The oral accounts illustrate how Líl̓wat traditions contain important details of Holocene landscape history that not only anticipated scientific findings, but also extend them, opening the door to productive synergies between Indigenous landscape knowledge and Western science.

Two Qw̓elqw̓elústen story threads are intertwined below, both of them observational and interpretive: one geological and the other preserved by Líl̓wat oral traditions. Both have published records, but part of the Líl̓wat account was the first to be published. The pathways leading to their intersection began independently, when ethnographer James Teit, of Spence's Bridge, undertook in 1898 and 1899 to record Líl̓wat cultural ways and oral history (Teit 1906, 1912; Boas 1906; Wickwire 2019); and when Edward Burwash, a geologist from Columbian College, Vancouver, spent several days in 1913 at Mount Garibaldi and determined it to be a volcano of Pleistocene age (Burwash 1914a, 1914b, 1918). Burwash realized that he had extended the known length of the Cascade Volcanic Arc into Canada and recognized that Mount Garibaldi eruptive products were the result of interaction with glacial ice. While he remained unaware of other volcanoes to the north, this laid the basis for such an expectation. Teit, whose travels included visits to the Pemberton area (Fig. 1A), was told a Líl̓wat story about the damming of the Lillooet River, followed by its breaching and a giant flood that impacted the river’s salmon runs (Teit 1912); but he did not know that the dam had been created by a volcanic eruption. In a footnote by Teit’s mentor and editor, Franz Boas, the story was attached to “the remote sources of the Lillooet River, in the main Cascade Mountains” (Teit 1912, p. 304), the name used at that time to include non-volcanic mountains now called the Coast Mountains, north of the Fraser River. Others would gradually draw these threads together.

Qw̓elqw̓elústen (“Mount Meager” of many authors) dominates the south side of upper Lillooet River valley, upstream from Pemberton near the head of the Lillooet–Harrison watershed (Bouchard and Kennedy 1977; Slaymaker et al. 2017) (Fig. 1A). It erupted ∼2360 cal year B.P. (B.P. = before A.D.1950) from a northeast-flank vent on Plinth Peak with flank collapses, a sub-Plinian eruptive column, a pyroclastic density current that temporarily dammed the Lillooet River, an outburst flood of the temporary lake, and the Bridge River (BR) ash–tephra that spread in a narrow fan as far eastward as central Alberta (Hickson et al. 1999; Andrews et al. 2014b; Jensen et al. 2019; Warwick et al. 2022). The event date is constrained by radiocarbon dates from wood (including tree stumps) within eruptive deposits at the volcano (Clague et al. 1995), wood in deposits of the temporary lake and debris flows (Simpson et al. 2006; Friele et al. 2008), organic materials above and below BR tephra in lake and bog cores and in alluvial overbank sections (Nasmith et al. 1967; Westgate and Dreimanis 1967; Mathewes and Westgate 1980; Vivian et al. 2017), and a varve-count from glacier-fed Hector Lake, Alberta, including the tephra (Leonard 1995).

Our discussion below cites Indigenous words from multiple sources that used differing orthographies, and we follow the original format if quoted. Group name formats are consistent with their websites. In Ucwalmícwts, the Líl̓wat7úl first language, the character “7” denotes a glottal stop or unvoiced throat closure: e.g., if an English-speaker stated the word “Scottish” with the “tt” silenced, it would be written as “Sco7ish.” Characters with an apostrophe above (e.g., l̓ and w̓   or l’ and w’ in some formats) indicate throat involvement modifying but not stopping the sound. An acute accent indicates dynamic stress of the syllable, sometimes indicated by capital letters (Bouchard and Kennedy 1977; van Eijk 2013; Lillooet Elders and van Eijk 2015). Initial capitalization of names is an externally imposed style and sources vary as to usage; e.g., for directly transcribed stories (Kennedy and Bouchard 2010) or for some group names on their websites.

In Ucwalmícwts, Qw̓elqw̓elústen (Qual-qual-OSH-tin; Bouchard and Kennedy 1977) means “cooked face place” or “cooked fire place”, as qw̓el means “done, ripe, or cooked”, and qw̓elqw̓el is a repetition, meaning “very cooked”; while the suffix -us is used for “face,” “hill,” or “fire” (also suggesting “personality” or “spirit”), and the suffix -tn indicates a location or place (also instrument or thing) (Abraham 2000). Heat, cooking, hot springs, and landscape changes recur as key elements in Líl̓wat oral traditions about this massif.

The Líl̓wat7úl, a distinct self-governing First Nation within the St’át'imc (Stl'atl'imx) Nations, are the southern of two dialect groups referred to in the past as “Lillooet”, with Líl̓wat also called Mount Currie (or Lower/Southern) Lillooet (Bouchard and Kennedy 1977) or Liluet-ō’l of Teit (1906). Líl̓wat lands extend diagonally northwest-southeast from west of Qw̓elqw̓elústen through to Lillooet Lake in an elongate zone straddling the upper Lillooet–Harrison watershed and crossing divides to the headwaters of rivers that flow beyond, west and southwest, to inlets of the Salish Sea (Fig. 1A). Downstream in the main watershed are the Lower St'latl'imx Nations: Samáhqwam, Sqátin (Skookumchuck), and Xáxtsa7 (Port Douglas; at upper Harrison Lake), also speaking Ucwalmícwts. Beyond, on Harrison Lake and River are the Sts'ailes (Chehalis) people, speakers of Upriver Halkomelem. Other Halkomelem-speaking Coast Salish peoples reside upstream and downstream of the Harrison River mouth on the Fraser River delta. Neighbouring First Nations to the west of the Líl̓wat include Skwxwú7mesh Úxwumixw (Squamish), shíshálh (Sechelt), and K’ómoks (Comox); and beyond the St’át'imc to the north and east are the Nlaka'pamux (Thompson) Nation (Teit 1906, 1912; Duff 1964; Bouchard and Kennedy 1977; York et al. 1993; Carlson 2001). Some of these peoples probably also possess oral traditions relatable to the eruption and the outburst flood; however, for succinctness, we only consider Líl̓wat examples to signal the need for further collaborative inquiry.

The Copper Canoe story, a partly metaphoric Líl̓wat oral narrative record from the Qw̓elqw̓elústen eruption through its resulting Lillooet valley outburst flood and failure of salmon runs, was told to James Teit, likely during visits to the Pemberton area in 1898–1899 (Teit 1912; Wickwire 2019), and was published well before the Western scientific accounts that echo it. Teit's prime source for “Lillooet” (Stl'atl'imx) knowledge was Chief James Stager (Qestíts̓a7) of the Pemberton Band (Kennedy and Bouchard 2010, p. 89), so it is possible that he told Teit this story. Parallels between this story’s and geological event-sequences show that it embodies direct landscape observations of ancient events. Other oral traditions discussed below suggest an eruptive plume, rockfalls, fumaroles, and possible explosive activity. Story elements once dismissed as fanciful are reconsidered here as mnemonic, as metaphoric characterizations that enhanced fidelity and transmissibility of unwritten landscape history across generations.

Mnemonic players include the Transformers, mythic beings who travelled ancient landscapes correcting things that had not been created properly, transforming them to modern forms. Landscapes of their time were by definition different from those of today and the Transformers had greater-than-human capabilities; yet Transformer stories reference specific modern landforms, and dramatic postglacial changes did occur in the Líl̓wat landscape. The Transformers properly empowered the land; therefore, later occupants who behaved in appropriate ways could receive these powers to varying degrees. Certain places, such as Qw̓elqw̓elústen, became residences of very strong powers. From the start people developed oral traditions about landscape, so over time “old stories” surely came to differ dramatically from the modern world. Amplifying this difference was the need for parsimony to compress an increasing mass of important information and allow its efficient transmission.

First Nations in this region were ravaged by Post-Contact period pandemics, with cumulative population losses reaching 90% to 95%. Missionaries and colonial governments further disrupted oral story transmission by removing or destroying sacred mnemonic artifacts, and by removing children to residential schools and preventing them from learning their languages (Duff 1964; Harkin 1994; Harris 1998; Boyd 1999). Despite these catastrophes, surviving and carefully curated oral traditions do provide linkages to times before colonization (Kii7iljuus and Harris 2005). Traditional environmental knowledge (TEK) is increasingly incorporated in environmental-change studies and resource management (Brush 1993; DeShield 1995; Cruikshank 2001; Tsuji and Ho 2002; Mazzocchi 2006; McDowell et al. 2023). Cruikshank (2012) cautioned that some scholars have incorporated TEK simply as data or evidence to corroborate other findings within established paradigms, rather than as knowledge or theory that could test and advance academic studies.

Central to our discussion is widespread (albeit not universal or monolithic) Indigenous belief in landscapes as sentient, capable of acting with intention. Features such as volcanoes and rivers can be portrayed as beings, and people can enter into relationships with landscape spirits for the common good (Bouchard and Kennedy 1977; Cruikshank 2001, 2012; McMillan and Hutchinson 2002; Barber and Barber 2004; Heyd 2008; Jones 2011; Scalise Sugiyama 2017; Latimer 2020). Pacific Northwest oral traditions indicate long-held knowledge of volcanoes and high montane settings, often focal to territorial identities. Skwxwú7mesh archaeologist Reimer/Yumks (2018) stresses the centrality of associated mountain practices: relations between people and volcanoes are more complex than is typically acknowledged by archaeologists. He highlights the social importance of volcanoes: people affirm social relations with focal peaks and respect them as spiritual beings, recalled and reinforced through narratives. Individuals must overtly introduce themselves to a mountain before ascending.

Volcanoes provide important tool-making materials, notably obsidian. While many archaeologists discuss lithic sources, tools also carry symbolic value as pieces of places, carrying the message that they come from places that “are significant parts of nature that are not separate from Indigenous cultures; they are respected and feared and offer physical, spiritual, and emotional strength for those who know them.” Thunderbirds and lightning are directly implicated in the making of obsidian (Reimer/Yumks 2018, pp. 5, 13).

Respect for mountains is widespread globally: they embody verticality, an axis reaching toward the above-world, making them loci and conduits of its powers (Barber and Barber 2004; Wilson 2005). Subalpine and alpine settings have long figured regularly in Pacific Northwest Indigenous resource procurement and ritual activities, attested by abundant archaeological and oral historical evidence (Franck 2000; Reimer/Yumks 2003, 2011, 2018; Mierendorf and Foit 2018; Inkpen et al. 2023).

The MMVC peaks (Fig. 1B) comprise Polychrome Ridge, Devastator Peak, Pylon Peak, Mount Job, Capricorn Mountain, Mount Meager, Plinth Peak, and smaller features, most (re)named formally between 1932 and 1980. Early settlers called the massif “Cathedral” (Robertson 1911, map; Ronayne 1971) but this duplicated two other British Columbia mountains (one 20 km north of Vancouver), so in 1924 it became “Mount Meager” from Meager Creek, named after settler J.B. Meager. Individual peaks of the massif were named in 1932 by alpinists, who in a two-week exploration made multiple "first ascents" and in the process confirmed the volcanic nature of the peaks (Carter 1932). “Mount Meager” therefore became restricted to one peak, but Qw̓elqw̓elústen signifies the massif.

The MMVC, still partly glacier-capped, formed during ∼2 Ma of eruptions from multiple centers to a complex-composite edifice with felsic through intermediate (dacitic and andesitic) to mafic lavas (Hickson 1994; Hickson et al. 1999; Friele et al. 2005; Simpson et al. 2006; Andrews et al. 2014b; Warwick et al. 2019, 2022; Morison and Hickson 2023). Multiple postglacial landslides and debris flows underscore continuing instability of this edifice, built of lava domes and pyroclastic eruptive material, often loosely stacked and vulnerable to hydrothermal pressurization (Read 1990; Friele and Clague 2004; Friele et al. 2005; Simpson et al. 2006; Siebert et al. 2010; Andrews et al. 2014a; Roberti et al. 2021; Morison and Hickson 2023). The glacier-fed upper Lillooet River is already sediment-loaded, so addition of mass-wasting debris has brought significant changes (Bovis and Jakob 2000). As a result, its delta into Lillooet Lake prograded 50 km down-valley in ∼12 000 years: this formerly 75 km lake shrank, through infilling, to its modern 25 km length within the timespan of Indigenous occupation (Friele et al. 2005; Friele and Clague 2009; Slaymaker et al. 2017) (Fig. 2).

The MMVC edifice was formerly higher but in the latest Pleistocene was covered, loaded/unloaded, and sculpted by the coalescent Cordilleran Ice Sheet (Clague and Ward 2011; Wilson et al. 2019); and erupted beneath or next to ice lobes (Andrews et al. 2014b; Russell et al. 2021). Mountains in this region were further sculpted by broadly synchronous, sequent Holocene (Neoglacial) advances occurring variably within these approximate limits: 8590–8180, 7360–6450, 4400–3970, 3540–2770, 1710–1300 cal B.P., and within the past millennium (Little Ice Age; LIA) (Reyes and Clague 2004; Reyes et al. 2006; Allen and Smith 2007; Menounos et al. 2009). The last was typically most extensive but the ∼2360 cal B.P. Plinth Peak eruption postdated a comparably strong advance/retreat cycle.

Dangerous landslides continue at Qw̓elqw̓elústen, reflecting fragmental volcanic deposits, high saturation, Neoglacial presence/absence history, and possible earthquake triggers (Evans 1992; Bovis and Jakob 2000; Friele and Clague 2004; Friele et al. 2008; Guthrie et al. 2012) (Fig. 2). An estimated 25%–75% of the total ∼10 km3 of Lillooet valley Holocene sediments came from the MMVC, mostly through repeated mass wasting (Simpson et al. 2006; Friele and Clague 2009).

Can oral traditions tell people about ancient “real-world” events, such as the Qw̓elqw̓elústen eruption and ensuing flood? Indigenous oral traditions range, globally and locally, from historical to supernatural narratives and from moralistic parables to entertainment, speaking multiple simultaneous messages at different levels of understanding. Simultaneously, they can portray the past and serve to guide and help negotiate ongoing social and power relations. It is presumptive to label a myth “fictional” when it can contain important embedded details from direct experience, especially of landscape (McMillan and Hutchinson 2002; Reimer/Yumks 2003; Barber and Barber 2004; Scalise Sugiyama 2017; Latimer 2020).

Oral traditions can survive across many generations, even millennia, but to do so must be economical (parsimonious) by (1) using metaphors to compress meanings, encapsulating complex ideas, (2) omitting some information the audience is expected to know already, and (3) using associative memory (mnemonic) devices (Budhwa 2002, 2018; Barber and Barber 2004). Landscape is a particularly effective mnemonic device, with story plot, players, and details attached to places to codify meaning and history (Morphy 1995; Basso 1996a, 1996b). Named features carrying attached messages populate empowered “landscapes of signification” becoming a class of material culture, evoking behavior and reinforcing group identity (Liebmann 2017; Wilson 2005, 2019).

Sequent places on trails (itineraries) can serve to order story elements (Basso 1996a, 1996b), and time’s passage may be encoded in distances (Morphy 1995). This is a very widespread phenomenon, whether a story-cycle from the Time of the Transformers (McMillan 1999; Thom 2003a, 2003b; Kii7iljuus and Harris 2005; McKechnie 2015; Latimer 2020), a cycle of Australian dreaming tracks (Morphy 1995; Kerwin 2010; Yunkaporta 2021), or an inter-island story itinerary in the Trobriands, Papua New Guinea (Harwood 1976).

Two millennia ago, Roman scholar Cicero lauded gifted orators who associated mental images with chains of localities to encode story sequence and content, and performed without written notes. This technique had come from people relying upon oral traditions. Harwood (1976), citing Cicero, refuted dismissals of myths as repeatedly reorganized, lacking fidelity, and showed that tying a myth to an itinerary of places, linked by a story plot, can insulate it across many generations from reorganization.

But landscape is more than a device for transmitting information: its form, processes, and history are integral to the message (Morphy 1995; Chiblow and Meighan 2022). People relying upon landscape resources must teach topographic knowledge effectively across generations. Hence, landscape is strongly represented in oral traditions (Barber and Barber 2004; Kennedy and Bouchard 2010; Scalise Sugiyama 2017) and place names (toponyms) are often descriptive, “crisp mental pictures” designed to inform (Basso 1996a).

The sentient-landscape view and presence of beings with special powers are evocative ways to embed landscape-history knowledge to teach people how the land behaves and needs respect. The presence of beings allows a historical or topographic sequence to be structured into a plot involving their actions and reactions, protecting the sequence. A named being may be ascribed a landform's powers and actions, playing a recurrent role (Cruikshank 2001; McMillan and Hutchinson 2002). We caution that this binding of events in sequence across generations is a function played by beings as story elements. “Function” in this context does not equal “purpose” because the beings and the stories simultaneously embody multiple culturally significant messages.

Indigenous peoples the world over have long sought explanations for landscape observations, with varied explanations but paralleling Western scientists in seeking to understand and predict landscape behavior. The “otherness” of many Indigenous oral traditions to Western observers reflects the sentient-landscape viewpoint. Western observers have viewed the physical environment as a container used by sentient beings, who hold dominion over it. Intersecting narratives can reveal productive overlaps between “myth” and Western science (Cruikshank 2001, 2012; McMillan and Hutchinson 2002; Barber and Barber 2004; McDowell et al. 2023).

A Klamath story from Oregon details Mount Mazama's destruction as a result of a furious battle between the Chief of the Above World (at Mount Shasta) and the Chief of the Below World (at Mount Mazama) for the love of a beautiful maiden. The contest comprised a series of events involving fire, burning ashes, and throwing of red-hot rocks, until Mount Mazama exploded and subsided into the earth, becoming Crater Lake. A soldier from Fort Klamath recorded the account in 1865 from a Klamath Chief who had learned it from his father (Clark 1953; Budhwa 2002, 2018; Barber and Barber 2004). Its detailed event sequence matches what was later determined by geologists, from 1886 onward, about the caldera-forming eruption ∼7600 cal year ago (Egan et al. 2015). A northern British Columbia Gitxsan ada'ok (“true traditions”) story about a giant bear that ripped its way down a creek, throwing trees aside and apparently damming Seeley Lake, strongly resembles geological evidence for a debris flow down Chicago Creek ∼3500 year ago (Gottesfeld et al. 1991; McMillan and Hutchinson 2002).

Líl̓wat oral traditions include much evidence of volcano- and valley-related geomorphic changes of times past. Líl̓wat  storytellers group oral traditions into two main types: (i) “ancient stories” or “myths” (in Ucwalmícwts, sptakwlth), involving Creator/Transformers and/or other beings with supernatural powers; and (ii) historical narratives (sqwéqweĺ); yet do not see this as a firm distinction, holding that events of ancient stories such as flood accounts did indeed occur (Bouchard and Kennedy 1977; Hanna and Henry 1995; Lillooet Elders and van Eijk 2015). Transformer stories in the Pacific Northwest are foundation myths, reciting itineraries by which these beings passed through the land transforming and naming imperfect places to create the modern landscape, at times even changing people into landforms (McMillan 1999; Kennedy and Bouchard 2010).

Performed in word and gesture, sptakwlth stories have protocols as to ownership, performance, and the required audience behavior; and can establish ethical/legal codes (Fiske 1998; Thom 2003a, 2003b; Lillooet Elders and van Eijk 2015; St’át'imc Chiefs Council 2018). Voicing and pauses can carry meaning, so transcribed words understate the full message. Historical and landscape details might have been introduced in retellings, but Transformer stories necessarily reference modern landscape features as resulting from ancient events.

Sqwéqweĺ stories, more narrative-historical and not as protocol-bound, tell of past times in the transformed world. Their enduring value can sometimes be demonstrated directly against other evidence, as with Wickwire’s (1994) correlation of details between Nlaka'pamux stories and Simon Fraser's expedition journal.

Líl̓wat traditions of Qw̓elqw̓elústen embody both sptakwlth and sqwéqweĺ elements, being a record of historical events framed as occurring in times of environmental transformation. Here we provide a geological narrative of the ∼2360 cal B.P. eruption and its aftermath; and for comparison, we follow it with elements of the Líl̓wat Copper Canoe story and other accounts to show their areas of striking similarity.

Qw̓elqw̓elústen (MMVC) peaks form a rough C-shape, open northward around a broad amphitheatre enclosing Job Glacier. An arcuate scar on Plinth Peak’s northeast face contains the ∼2360 cal B.P. eruptive vent (Figs. 1B and 3B). Job Glacier’s ongoing retreat revealed fumaroles in about 2015, creating glaciovolcanic caves (Warwick et al. 2019; Roberti et al. 2021; Unnsteinsson 2022); and the amphitheatre documents a major Holocene landslide related to a 2 km long debris deposit at the Job Creek–Lillooet River confluence. This appears linked to a valley-wide debris flow (volcanic diamicton) documented by sediment coring to extend ∼50 km downstream in Lillooet valley and dating at most a few centuries before the eruption. The cores also indicate an overlying hyperconcentrated flow deposit (i.e., reflecting a water–sediment slurry) that likely represents an outburst flood (Friele et al. 2005, 2008; Simpson et al. 2006). Slide-related decompression of underlying materials could have perturbed the MMVC stress field, mobilizing volatiles and hastening magma ascent (Roberti et al. 2021). Much of the ice present today accumulated after the eruption, especially during the LIA.

Deposits indicate an initial sub-Plinian eruption: violent and explosive, with a strong upward component and a lateral component. The column likely rose 14–18 km and was deflected eastward with the jet stream, carrying BR ash–tephra eastward to central Alberta (Westgate and Dreimanis 1967; Mathewes and Westgate 1980; Clague et al. 1995; Jensen et al. 2019). A less explosive Vulcanian phase ensued, then a short lava flow at Fall Creek. The corresponding eruptive assemblage comprises (i) pyroclastic airfall from five closely spaced eruptive phases, each evolving from phreatomagmatic to magmatic; (ii) welded to non-welded, pyroclastic block-and-ash flow with entrained charred logs and pumice blocks; and (iii) dacitic lavas (Stasiuk and Russell 1990; Stasiuk et al. 1996; Stewart et al. 2002; Michol et al. 2008; Andrews et al. 2014a, 2014b; Morison and Hickson 2023).

A largely impermeable, welded pyroclastic ∼110 m high dam of lava and block-and-ash flows capped by avalanche deposits temporarily blocked the Lillooet River, damming it and forming a lake upstream from what are now Keyhole Falls (Fig. 3). The lake's ∼60 m rise led to catastrophic failure of the less resistant upper part of the dam weeks later, sending a major outburst flood (lahar) at least 65 km down Lillooet valley. A spectacular result is the still-present narrow 2.5 km slot canyon, 60 m deep, cut in a matter of hours into the lower, more resistant part of the dam while it was still cooling, and feeding Keyhole Falls (Read 1990; Michol et al. 2008; Andrews et al. 2014a) (Fig. 3). Irregular, fan-shaped columnar joint-fracture patterns are visible from the dam’s cooling (Fig. 3A). The outburst scoured a wider, deeper, ∼2.5 km canyon below the falls and carried house-sized welded-breccia blocks for several kilometres; and continued beyond to Lillooet Lake, adding to its progradational infill (Read 1990; Friele and Clague 2004; Friele et al. 2005; Simpson et al. 2006; Slaymaker et al. 2017).

In flood, the river still entrains frothy pumice boulders, floating like corks, “a startling sight to strangers unaccustomed to seeing boulders floating around” (Ronayne 1971, pp. 19–20). Pumice is mined on the bank near Salal Creek, upstream from Keyhole Falls (Fig. 3B). The proximal eruptive deposits define a ∼N63° E axis (Hickson et al. 1999; Andrews et al. 2014b).

It was distal tephra from the ∼2360 cal B.P. eruption that led gradually to discovery of the Plinth Peak vent and the mountain’s volcanic heritage. Fresh-looking white pumice was first recorded north of the Lillooet drainage (Bateman 1914a, 1914b) during early 20th-century bedrock mining that built on clues from a brief 1859 placer gold rush (Robertson 1911; Ronayne 1971; Decker et al. 1978; Duclos and Duclos 1995). The BR and Lillooet mining districts widely exhibit superficial tephra on valley floors, hillslopes, and summits (Bateman 1914b, p.193; McCann 1922; Cairnes 1937). Bateman hypothesized volcanic vents between the Bridge and Lillooet headwaters, possibly linked to Lillooet valley thermal springs that had been noted by Robertson (1911) at “Cathedral Mt.” in a rough sketch map of what was then a little-known area. Stevenson (1947), describing Bralorne Mine tephra exposures on Hurley River, quoted a 1946 personal communication from J.M. Cummings (B.C. Department of Mines), who had found in the Lillooet River headwaters that Mount Meager was at least a local source of pumice. Cummings cautioned, “It is debatable whether Meagher [sic] Mountain was the source of the Bridge River pumice or merely one of a number of local sources” (Stevenson 1947, p. 548).

Nasmith et al. (1967) narrowed the source to Plinth Peak where ash and pumice were thickest. The vent location was further refined to the slope scar on Plinth Peak’s northeast flank, obscured by headwall collapse (Fig. 3B) (Read 1990; Andrews et al. 2014b).

Early prospectors skirted the MMVC area (Robertson 1911). Some Líl̓wat people later joined in minor prospecting well downstream from Qw̓elqw̓elústen; Chief Paul Dick (Mount Currie) prospected with others near Tenas [Little Lillooet] Lake into the 1930s (Decker et al. 1978). Mount Currie elder Charlie Mack heard that before 1924, a Tsilhqot'in (Chilcotin) hunting party found gold at the Lillooet River headwaters and “used it for shot to kill groundhogs [marmots]” (Kennedy and Bouchard 2010, p. 41). In 1932, three prospectors searched the Lillooet Glacier area, “skirting Meager Mountain” and “rode the length of Polychrome Ridge, but staked no claims… the mountain's great pumice deposits were of no interest” (Decker et al. 1978, pp. 163–164).

The Transformers

Líl̓wat oral traditions about Qw̓elqw̓elústen include some from the mythic cycle of the Transformers, others involving acquisition of spiritual powers vested in the landscape, and others of a more narrative historical character. The founding Transformers (ats’íma7lh) were three (or four) brothers, their sister (transformed into their canoe), and another being (Teit 1912; Kennedy and Bouchard 2010). Stories of their activities provide the framework of Líl̓wat spatial logic, as detailed by Kennedy and Bouchard (2010): the Transformers put the world right by using their powers of transformation to modify the landscape into its contemporary forms. This knowledge is therefore associated with the earliest times of occupation of Líl̓wat territory and with those who founded it and their laws (Teit 1906, 1912; Bouchard and Kennedy 1977, pp. 13–17; St’át'imc Chiefs Council 2018).

The Transformers paddled up the lower Fraser River from the coast, entered Harrison River, and then went up Harrison Lake, lower Lillooet River to Lillooet Lake, and the upper Lillooet River to its source near Qw̓elqw̓elústen at Lillooet Glacier, establishing specific named places. They then traveled through the Anderson and Seton Lakes areas to Cayoosh and Lillooet, before returning westward to Mount Currie. Upon completing this, they returned via Pemberton and Green Lake to Skwxwú7mesh (Squamish) territory and the coast (Teit 1912; Bouchard and Kennedy 1977; St’át'imc Chiefs Council 2018; Kennedy and Bouchard 2010).

In an interview with Angelbeck and Jones about the upper reaches of Líl̓wat7úl traditional territory, Líl̓wat Elder Harry Dick related that the Transformers stood atop Qw̓elqw̓elústen long ago. He recounted that the Transformers could see the ocean to the west, for the mountain was higher then, “before it blew its top,” the highest mountain in this territory (Angelbeck et al. 2011, p. 261).

The Copper Canoe story

Another story tied directly to Qw̓elqw̓elústen was recorded by Teit (1912, in edited form) and a more detailed version was recounted in the 1980s by Líl̓wat Elder Charlie Mack (b. 1899–d.1990) (Kennedy and Bouchard 2010). The Copper Canoe story begins with a quest for spiritual powers (áx7a) at or near Qw̓elqw̓elústen, recognized since earliest times as an especially powerful place. Its remoteness, well northwest of Líl̓wat villages near Pemberton, also suited it for individual quests. In the story, two shaman brothers (Ucwalmícwts, scwená7em) sought training near the mountain’s base, as it was hi7 (spiritually powerful). They acquired certain transformative powers, and their itinerary through Líl̓wat territory began in the opposite direction from that of the founding Transformers.

The brothers travelled in a Copper Canoe that had special powers. Northwest Coast Nations for 2000 or more years made artifacts from cold-hammered native copper, viewed as spiritually powerful, animate, and associated with supernatural beings; so a more substantial Copper Canoe is a familiar story motif (Teit 1912; Kennedy and Bouchard 2010; Cooper et al. 2020). Teit's (1912) record of this story spoke only of a canoe (clearly with considerable powers) but another story in his compilation (“The Gambler”) included a Copper Canoe, and another (“The Poor Man, or The Origin of Copper”) told of a man fishing a large piece of copper from a lake.

The Copper Canoe story as told by Charlie Mack is presented and discussed in detail by Kennedy and Bouchard (2010, pp. 19–28); and here we focus upon portions that relate to Qw̓elqw̓elústen and the upper Lillooet valley to reveal an eruption-flood linkage not previously documented. We acknowledge that the story itself is owned by the Líl̓wat Nation and limit our discussion to selections from publicly available source material: our discussion is not intended as a new “performance” of the story.

The brothers camped near the mouth of Salal Creek, at the northeast footslope of Qw̓elqw̓elústen (Fig. 3B) (Kennedy and Bouchard 2010, pp. 20–21). Teit (1912) was less specific, but clearly indicated the same general area of the upper Lillooet River (Líl̓watátkwa):

“Two brothers lived at the very head waters of the Upper Lillooet River, and spent most of their time training themselves in the neighbouring mountains for they wished to become great” (Teit 1912, pp. 303–304).

Mack recounted that one of the scwená7em had fallen sick. His brother had a vision of a Copper Canoe, took hold of it from his dreamworld, and pulled it into reality:

“And they were bathing in there, singing at night, dreaming. They said they were scared sometimes when they dream something bad. If they are big enough, they grab a hold of it, what they dream. And he caught it. That is how they got this canoe…. they went to sleep and dream about a canoe. A canoe, a Copper Canoe, and he grab a hold of it. When he got it, well, he own it; the canoe didn't disappear…. It start above the canyon, above the waterfall” (Charlie Mack, quoted in Kennedy and Bouchard 2010, pp. 20–21).

The brothers jumped in and floated down Líl̓watátkwa, shaping its course, further altering the landscape. Their first obstacle was a wall of stone at the base of Qw̓elqw̓elústen, holding back a lake: “Well, that water was tall there, over 200 feet [60 m] high before” (Kennedy and Bouchard 2010). Teit (1912) had recorded,

“They came to a place called Ilamüx. Here there was a rock which dammed the river. They made a hole through it to allow their canoe to pass. Even at the present day it appears like a stone bridge across the river” (Teit 1912, pp. 303).

Mack spoke of the place smúm'leq [Múm̓leq] in such a way as to suggest it was the dam and the associated rapids, stating that it had been created by bad persons:

“Well, some bad persons on the way down, you know, the Copper Canoe stop it, and tame it. Like that smúm'leq used to be rougher than hell, up to the sky and high” (Charlie Mack, quoted in Kennedy and Bouchard 2010, pp. 21).

Kennedy and Bouchard (2010, pp. 24–25) stated the location of smúm'leq as “above Pebble Creek, where the rock comes down to the river,” which might collectively signify the long slot-canyon rapids and (as Múm̓leq) its modern Líl'wat usage for Keyhole Falls and the swirling pool below. With their Copper Canoe, the brothers sliced through the rock wall—“cut right through there,” said Mack (Kennedy and Bouchard 2010)—forming the slot-canyon from which the water now falls into this pool (Figs. 3 and 4).

Given the changes wrought by the eruption and damming event, it is expectable that pre-eruption toponyms might have taken on modified meanings or that new toponyms appeared. The name Sq'em'p [Sq̓em̓p] has been used for “the blockage in the canyon” (Kennedy and Bouchard 2010, pp. 24; Jones 2011), but also denotes a creek upriver from Salal Creek (Angelbeck et al. 2011, pp. 40–41); so perhaps at some time this name embraced a larger area. Teit's “Ilamüx” is a different name that appears to have been applied directly to the rock-dam. The name “Keyhole Falls” was given by settlers because of the narrow cut’s shape, wider at the base (Fig. 3A). Conceivably, the large landslide event down Job Creek valley to the Lillooet, at most a few centuries before the eruption and not far upstream from it, could also have fit the description of rock that caused “a blockage in the canyon.”

Continuing downstream, the brothers encountered many other still-recognizable landmarks and several obstacles, detailed by Kennedy and Bouchard (2010). One such landmark is the face-to-face meeting of two large, sometimes “angrily” competing tributaries (North Creek and South Creek) from opposite sides of the Lillooet River (Fig. 4). Obstacles included mountain walls blocking the brothers’ path, high gravel beds, knots of bulrushes, and areas of particularly rough water. Blocking mountains could refer to sharp-angled river bends causing flood run-up where the valley is narrow. With the Copper Canoe, they surpassed these obstacles as they continued down the Lillooet–Harrison system’s rivers and lakes to the Fraser River, then onward to the Salish Sea, the homeland of the salmon (Fig. 4).

However, the Lillooet River salmon were blocked by a massive fish weir. This is a recurrent theme in Pacific Northwest narratives, many referring to political conflict (Ritchie and Angelbeck 2020). It is worth noting, however, that debris-flow outburst floods throughout this mountainous region entrain riparian trees and can transport them long distances, causing downstream log-jams. These may temporarily block fish migration but once breached can also improve fish habitat by creating persistent pools (Montgomery et al. 2003); hence, they could be linked conceptually and metaphorically to fish traps. The brothers ingratiated themselves with those who maintained the weir and pulled some stakes from it, releasing the salmon upriver. One of the brothers “hollered up the river, “Fix up a smokehouse, the fish is coming up now””  (Kennedy and Bouchard 2010, p. 27). In so doing, the two scwená7em, using powers earned at the base of Qw̓elqw̓elústen, (re)introduced salmon into Líl̓wat territory: a key element of the narrative. The brothers accompanied the fish upriver all the way back to Qw̓elqw̓elústen, where they cooked salmon in waters of one of the hot springs (thus, waters hot enough to cook fish), according to Louie Joe (Angelbeck et al. 2011). In Teit’s version, they “returned to their home at the head of the Upper Lillooet River, and they made near their house the hot springs called Tcîq [Tsek], which they used for cooking their food” (Teit 1912, p. 304). The brothers thereby completed the cycle, returning to the starting point, symbolically binding the powers of Qw̓elqw̓elústen to the salmon’s return.

In a related story, Baptiste Ritchie (Mount Currie) recounted that when people returned to their former lands in the upper Lillooet valley, they found the valley floor “bare because the water was so rough it took all the ground off the valley.” In contrast, so much material had been redeposited downriver that the Pemberton Meadows area had become livable. “All those swamps were filled and the [upper valley] people came down there instead of staying at their place… because the valley where they lived was sort of washed down, washed out” (Kennedy and Bouchard 2010, p. 98). At the time of the eruption and outburst flood, the delta at the head of Lillooet Lake was ∼15 km upstream from its present location, placing the lake’s head near modern Pemberton (Slaymaker et al. 2017) (Fig. 2). The debris-flow deposit extended a few km into the lake, beyond Pemberton.

Landscape changes of the Copper Canoe story and other accounts therefore parallel and complement geological evidence from Qw̓elqw̓elústen and along Lillooet River in the wake of the eruption; e.g., the valley-wide outburst flood that, like them, travelled in a frenzied way downstream, overriding obstacles in its way. Overall, the story recounts (1) something very powerful (“very bad”) at Qw̓elqw̓elústen, (2) a dam (wall) being created (by “bad persons”), backing up a 60 m deep lake, (3) a dramatic event that breached the dam and sharply cut a narrow slot-canyon, (4) release of a great volume of water, flooding the valley floor, (5) scouring of the upper valley and sediment transport to at least Pemberton Meadows, where it filled swamps and made the valley floor habitable for people, and (6) clogging (including log-jamming) of the river downstream, preventing salmon from ascending until the jam was breached. Both the story elements and their sequence match the geological findings (e.g., Andrews et al. 2014a); and the blockage of the river to salmon, displacement of people, and substrate changes at Pemberton Meadows add important new information. As in the story, thermal springs occur at the massif’s base: Pebble Creek/Keyhole Hot Springs, next to Lillooet River northeast of Plinth Peak, downstream from Keyhole Falls; and Meager Creek Hot Springs, south-southeast of Mount Meager (Hammerstrom and Brown 1977; Líl̓wat Nation 2021).

During their quest for power, the brothers had “dream[ed] something bad” just before their downstream odyssey. This and the “bad persons” who built the dam may reflect a storyteller's conscious effort to avoid directly naming the spiritually powerful eruptive event, as well as indicating something the audience was expected to know already.

The time ascribed to the Copper Canoe journey is thus established, because the brothers’ quest for power was at, and echoed in, the ∼2360 cal B.P. eruption and damming of Lillooet River. The founding Transformers were earlier: they had already transfigured the empowered landscape within which the brothers lived and sought power.

The forbidden mountain story

Ronayne's (1971) Pemberton valley local history includes a second-hand but important fragment of a Líl̓wat story concerning Cathedral Mountain. Although a colloquial English retelling, it has a mythic tone and indicates a code of respect:

“These Indians have a taboo on very high mountain peaks…. A huge mountain [elsewhere given as Cathedral] at the head of Pemberton Valley, of volcanic origin, with five black needles reaching into the sky, has another story. Here against the medicine man's orders, two daring young men climbed the forbidden mountain, reserved for some heathen gods, and for their sin were condemned forever to sit cooking their own heads over a ghostly campfire.” (Ronayne 1971, p. 93).

The story fragment contains the elements “fire,” “head,” “cooking,” and “place” (“here”): essentially, the name Qw̓elqw̓elústen. All stories suggest reticence to state this name directly, using “Meager” in recent tellings or metaphorically citing the three components. A “forever” campfire suggests a fumarole plume as recently reappeared at Job Glacier, having been hidden by LIA pulse ice; therefore, this story likely predates the LIA.

High country activities

Charlie Mack also stressed the economic importance of Qw̓elqw̓elústen for the Líl̓wat people. His family and friends camped yearly at the massif for two weeks, reaching it by canoe and hiking. Men hunted and women gathered edible roots. Hunters not only used bows and arrows, but also had long poles to vault between cliffs, to assist in climbing rubble, and to cause mountain goats to fall down unstable slopes (Bouchard and Kennedy 1977; Maud 1978; Kennedy and Bouchard 2010, pp. 106–108). Rockshelters large enough for hunting parties and other seasonal camps are located along the river at the mountain’s base (Sheppard et al. 2015).

The huge snake

These seasonal visits are a clue to the timing of a more narrative historical account involving a huge snake, recounted in 1968 by Karen (Pascal) Gabriel, of Mount Currie, in a Pemberton local history. This story, “About Long Ago,” appears to record an eruption:

“Then one day while the women were cooking for their children, all of a sudden they heard terrified cries coming from their men. They dropped the clay bowls which they were holding and ran to see what was happening. The sky darkened and they could see a huge snake cross from hill to hill. They ran into the caves they were living in and never came out until the snake had disappeared onto the next hill. The people began to pray to their gods, because they thought the gods were angry and had sent the snake after them. They began to worship any snake they saw afterwards” (Decker et al. 1978, p. 38).

Such a startling event with skies suddenly darkening suggests passage of an eruptive plume overhead; its shadow crossing the landscape. Its unprecedented nature suggests the initial sub-Plinian eruption. The “snake” is a metaphor enhancing memory: a “snake” crossing the land could be something narrow, moving in a sinuous way, much as a line of automobiles is often described as “snaking down the road.” A narrow shadow is consistent with ash–tephra distributional evidence that some of the plume was streamlined as prevailing westerlies carried it eastward. Hazard modelling indicates a broader local spread of the tephra fan, but this would also include tropospheric transport of tephra from episodic non-Plinian eruptions over several months (Warwick et al. 2022).

Karen Gabriel received the story from her grandmother, who had received it from her great-great grandmother. “Clay bowls” signal a time before Historic trade of metal vessels. Pottery was not used in this region, but pecked or carved argillite and soapstone mortars were used for food preparation and ritual purposes, were not readily portable, and would have been left behind by people rushing outside (Teit 1906; Reimer/Yumks et al. 2016). “Caves” were the excavated semisubterranean (“underground”) winter houses or s7ístken of Líl̓wat and neighbors (Teit 1912; Bouchard and Kennedy 1977). They were used from about November to February (late fall until tree leaves began to reappear); though a few people remained year-round (Teit 1906, 1912; Bouchard and Kennedy 1977, p. 64). Presence of men, women, and children in the story supports a time other than the late spring-to-fall salmon-harvesting period, when many people were away at fishing camps (DeShield 1995); and therefore a broadly “fall-to-winter” eruption when few people would have been at the mountain.

Rocks blown in the air

Charlie Mack described Qw̓elqw̓elústen as hi7, spiritually powerful, and dangerous, a “monster” (Kennedy and Bouchard 2010, p. 18). In a story from the 1900s a hunter named 25-Mile Jim along with James Stager and Tepayás, a shaman, tried to climb the mountain but wind gusts became too strong:

“Well, they blowed away…. They just walked off, they can't do nothing. They were hunting, they go up there to hunt goats and they couldn't make it. Then they start to make a noise; I guess they echo to the high mountain, and it start to blow. Well, 25-Mile Jim, he says: ‘I seen rocks go up in the air.’ It was blowing so hard, on the other end of the rock. It seems the echo started the hi7” (Mack, in Kennedy and Bouchard 2010, pp. 19, 106).

Given presence of fumaroles and abundant groundwater, minor phreatic (steam-burst) eruptions could have occurred at any time. Mack told this story in 1988 after he, with Kennedy and Bouchard in a helicopter, had approached to land on Qw̓elqw̓elústen (Kennedy and Bouchard 2010, pp.13–15). Having second thoughts, he told the pilot not to land. Tumbling rocks stirred up by propeller wash reminded him of the story, signaling it was not the right time or way to approach the mountain.

We know that there are other stories. Elders recall hearing, decades ago as children, stories from their grandparents about a “river of fire” and about a time when people walked in ashes, sometimes up to their knees (Angelbeck et al. 2022).

The Qw̓elqw̓elústen area constitutes a vital Líl̓wat heritage landscape, a focal concern because narratives weave it and Múm̓leq into multiple heritage aspects: Transformer stories establishing cultural priorities, a place of reverence for its spiritual powers related to the Copper Canoe brothers, hot springs for spiritual bathing, historical events including a territorial defense battle, traditional subsistence loci, and other sacred places (Angelbeck et al. 2011; Jones 2011). Qw̓elqw̓elústen is the origin landmark of the Copper Canoe journey that made Líl̓wat territory more closely into the landscape known today. The Líl̓wat Nation proactively protect their archaeological heritage (Angelbeck and Jones 2019); and the land use plan recognizes this area as an á7x7úlm'ecw, “spirited ground” (Líl’wat Nation 2008), enshrined and performed in myth and fundamental to the Líl̓wat worldview.

A memory landscape map created by Johnny Jones (2011) after the 2010 landslide illustrates traditions related to this powerful area (Fig. 4), building upon the style of pictographic art found at important locales in Líl̓wat territory and the surrounding region (Teit 1906; York et al. 1993; Arnett 2017). This map is in the colour of red ochre, a protective device associated with guardian spirits and possessing its own agency (Arnett 2017). Jones, as a Líl̓wat heritage-keeper, has composed several such maps, featured in roadside kiosks and an archaeological publication, to depict parts of Líl̓wat territory. The intention is to give visitors (outsiders) a first lesson in Líl̓wat spatial logic by integrating traditional pictographic, active forms within a cross-culturally recognizable geographic frame (Jones 2011). Qw̓elqw̓elústen is shown as a dangerous spirit, with a vortex-like spiral at its centre and a fan of snakes' heads with forked tongues (landslides) leading down its slopes. The major 2010 landslide along its Capricorn Creek flank (Fig. 2; Guthrie et al. 2012) is depicted with clawed limbs reaching further downslope along Meager Creek and Lillooet River. Whorls indicate the spiritually important hot springs on the northeast and southeast flanks.

Líl̓wat use of routes to the hunting grounds in the headwaters of west-flowing rivers is exemplified by a dashed trail marker past Lillooet Glacier (Fig. 4) and was in part linked to trade of mountain goat skins, hair, and blankets to neighbours for coastal goods (Teit 1906; Reimer/Yumks 2003; Kennedy and Bouchard 2010). Charlie Mack described Qw̓elqw̓elústen as the “place where mountain goat heads are cooked” (Kennedy and Bouchard 2010, p. 106), and Jones depicts mountain-goat hunters near Qw̓elqw̓elústen, one with a pole.

It is said that at one time people hunting or seeking power could jump across the river just above Keyhole Falls (Múm̓leq), where the vertical-walled Lillooet River slot-canyon narrowed to a few metres at the rims, widening below (Figs. 3A and 4) (Jones 2011). Harry Dick called the seekers scwená7em (shamans), individuals undergoing spiritual training. To fail in the jump meant a fatal fall to the river below, flowing to its Keyhole Falls outlet and dropping ∼35 m to eddying pools (Angelbeck et al. 2011; Jones 2011). Charlie Mack also stated that mountain-goat hunters could cross the canyon at the narrowest to ascend Mount Meager (Kennedy and Bouchard 2010). However, ongoing erosional retreat of the joint-fractured upper slot-canyon walls, (Fig. 3A) driven by such processes as freeze-thaw, thermal expansion–contraction, and earthquakes (e.g., the giant 1700 CE Cascadia subduction zone event) slowly increases the gap and the cliffs are now too treacherous for access (Andrews et al. 2014a).

Near the edge of the Keyhole Falls overlook is a rock shelter likely associated with scwená7em, because it is close to views of upper Lillooet valley from above the falls, as well as of the Qw̓elqw̓elústen peaks (Sheppard et al. 2015). Community member Tony James told of staying at the rockshelter during a training in quest of spirit powers (Angelbeck et al. 2011).

By contextualizing Líl̓wat volcano-related stories in their landscape settings, we conclude that, though some involve people with mythic powers, they directly record the eruption of Qw̓elqw̓elústen and the resulting landscape changes. The Copper Canoe story is the strongest thread woven into the fabric of this heritage landscape. Qw̓elqw̓elústen merges time and space as a locality in which “wisdom sits in places,” as Western Apache elder Dudley Patterson told Basso (1996a, 1996b). “Wisdom” in this sense entails a heightened mental awareness and steadiness to anticipate and avoid harmful events. It draws upon traditional landscape (geographical) and historical knowledge: stories encode a landscape of signification and the landscape in return evokes the stories and their lessons.

A century after the first publications of southwest British Columbia oral traditions, Maud (1982, p. 71) was still assessing stories in search of “masterpiece myths,” elegant fiction built around characters and plots. He commented that “Admittedly, in the worst tellings, these Transformer stories read just like one old transformed rock after another,” albeit conceding that the stories had “practical significance.” Some had been told in the simplified vocabulary of Chinook Jargon, the regional trading language (Bouchard and Kennedy 2002), and early published versions were stylistically filtered by their collectors or editors, potentially obscuring rather than celebrating their content. We argue instead that embedded landscape information had long served to establish places and sequential itineraries, crucial knowledge to be taught across generations (Scalise Sugiyama 2017; Latimer 2020).

Cruikshank (2001, 2012) called for studies of the “intersection between myth and science.” This is consilience, bringing together principles and information from diverse sources and viewpoints to build greater understanding. Indigenous traditional knowledge and Western science wear different contextual shells, but both embody observations and dedicated attempts to understand the world, to predict its actions. The scientific attitude of skepticism (testability) can lead to premature rejection of information in oral traditions. Once one factors in the Indigenous view that geomorphic processes represent beings capable of acting with intention in powerful, sentient landscapes, the stories reveal embedded landscape history, extending scientific knowledge and providing both observations and testable hypotheses.

The first published statement that the Lillooet River had been dammed by a rock feature that was then cut through, producing a narrow canyon, was that of the Líl̓wat storyteller whose words from 1898–1899 were reported by Teit (1912). Yet Teit’s published version focused more upon the characters and the plot, packaging the story as “The Salmon-Men; or The Origin of the Salmon,” and limiting the landscape details. The extent to which the 1912 version was selectively abridged by Teit’s patron and editor, Franz Boas, is unclear, but Boas was extremely interested in regional diffusion of widespread story motifs (Boas 1896, 1898; Wickwire 2019). Many of Teit’s voluminous field notes relating to published works, and some of his manuscripts, were destroyed by Boas after Teit’s death and others are still incompletely studied (Wickwire 2019, p. 268). It appears that Teit was not fully fluent in a St’át'imc dialect though he was fluent in Nlaka'pamux (Thompson), Secwépemc (Shuswap), Syilx (Okanagan), and the ubiquitous trading language, Chinook Jargon (Wickwire 2019, pp. 207, 267). Teit in 1898–1899 may have used an interpreter, but his 1906 monograph records that “Lillooet” (generalized St’át'imc) were particularly adept in using Chinook, a language with which he was familiar through his work as a trader and guide. Teit compiled word and toponym lists during his visits to Lillooet, Douglas, Skookumchuck, and Pemberton (Wickwire 2019, pp. 123–124; Boas 1906; Teit 1906), and would have made notes about the grammar. The preface to Teit’s 1912 compilation included credit to Homer Sargent as a sponsor (through Boas), something hidden from Teit until 1910 (Wickwire 2019, p. 135), but Wickwire does not record that Teit made yet another trip for Sargent at that time, to continue documenting Líl̓wat traditions. More likely, the report itself was facilitated through Sargent’s financial support and was based primarily upon his earlier field work. Boas’ endnotes accompanying Teit's 1906 monograph do indicate that Teit traveled through the “Lillooet” area again in 1903, at which time he made population estimates for inhabited villages (Teit 1906, p. 292).

It is no accident that oral traditions of landscape transformation pervade this region, which saw conspicuous and often sudden geomorphic changes within the span of human presence. Yet region-wide occurrence of Transformer stories led early ethnologists to emphasize regional “borrowing” over local detail. Boas was driven to demonstrate cultural diffusion of ideas from Asia to the New World around the Pacific Rim (Wickwire 2019), and for Nlaka'pamux (Thompson River Nation) stories, he wrote dismissively, “It appears that a considerable number of tales were borrowed bodily from the coast tribes, and were incorporated ready-made” (Boas 1898, p. 15). This minimized the fact of intermarriage between neighboring peoples: stories carried from one to another became the inherited and owned history of descendants in both groups. To the “borrowed” dismissal was added the venerable trope that every culture has a flood (deluge) myth. But there was not simply one archetypal flood theme: many Pacific Northwest stories document local consequences of real-world events, notably the Cascadia subduction zone earthquake, tsunami, and landslides of 1700 CE (Budhwa 2002, 2018; McMillan and Hutchinson 2002; Ludwin et al. 2005). Accordingly, some claimed “borrowings” of flood stories could instead record the well-documented underlying commonalities of debris flow and outburst-flood dynamics from drainage to drainage, as evident in the pattern of upstream scouring (reaming) of riparian corridors, downstream deposition of gravelly sands, and creation of major log-jams affecting fisheries, all (for example) accompanying the 1980 eruption of Mount St. Helens (Major and Mark 2006) and 19th-Century activity of Mount Baker (Tucker et al. 2007).

The Copper Canoe story tells of shaman brothers who obtained powers of transformation, yet were not the founding Transformers. The brothers sought and received powers within a land already transformed, at an already powerful, iconic locality: Qw̓elqw̓elústen. Their story and that of the more ancient founding Transformers show how oral historical narratives, to teach new generations, were streamlined by metaphor for efficient transmission, yet preserved integrity through use of mnemonic devices. These are not simply “stories” in the sense of self-contained, plot-driven packages with definable beginnings and ends. They carry multiple messages at different levels, essential for cultural continuity. Landscape, as a mnemonic device, is not only an organizational framework, but can also be a vital part of the message, affirming and describing the traditional territory to teach and preserve this vital information across generations. These lands are detailed in orderly fashion: an orally transmitted, annotated historical atlas with linked “pages” overlapping in content. Each storyteller can select what part of the geographic field to perform and which other theme(s) to emphasize, depending upon context and audience. Retelling of oral historical traditions, then, in part anticipated today’s use of Global Positioning System (GPS) devices: once coordinates (locations) are selected, the route is mapped out (performed). Some GPS devices speak to users, once again orally performing itineraries to guide behavior as in the days before writing. Increasingly, these devices are linked to sources that tell associated site-specific stories: a function that, while applauded as “new” and as “augmented reality,” has a very long pedigree in oral traditions.

We have come full circle. The Copper Canoe story not only anticipated, but also extends scientific knowledge by recording the return of the salmon fishery after its loss due to the pyroclastic damming and the associated outburst/debris-flow event and logjam. An associated story also records displacement of people from upstream of Pemberton Meadows, and that upon their return some occupied newly transformed valley-floor lands downstream from their previous, now-scoured lands. This opens new pathways for archaeological and paleoenvironmental investigation, as well as signaling the likelihood that similar fishery disruptions and human population displacements accompanied other ancient Holocene flank collapses and debris flows in this volcanically influenced drainage basin. The human costs of these ancient events have yet to be appreciated adequately in the literature. Oral traditions and geological history together affirm that similar events will happen again, and that this storied, powerful, and active landscape deserves enduring respect.

We thank Xzúmalus/Roxanne Joe, Project Coordinator, Lands and Resources; and Kik7ak/Councilor Helena Edmonds, of Mount Currie, for their advice about place names and transcription of Líl'wat words. Quotes from The Lil'wat World of Charlie Mack, © 2010 by D. Kennedy and R. Bouchard, Talonbooks, B.C., are reprinted by permission from the publisher and we thank Kevin Williams of Talonbooks for his guidance. We thank Charmaine Carpenter and Niki Madigan of the Pemberton & District Museum & Archives Society, Pemberton, B.C., for guidance and clearance to quote from Beyond Garibaldi (now out of print) and Pemberton: The History of a Settlement (still available from their museum). The map in Fig. 1 was adapted with permission from the online Líl'wat Nation (2006)Lil'wat Land Use Plan; and the accompanying Geological Survey of Canada photograph was modified with permission from Kirstie A. Simpson, after Simpson et al. (2006). Figure 2 was adapted from Slaymaker et al. (2017) with permissions from Olav Slaymaker, Pierre Friele, John J. Clague, and Springer International Publishing. We thank Steve Quane for permission to use his photograph of Keyhole Falls in Fig. 3A, and J. Kelly Russell for permission to adapt the photograph in Fig. 3B (after Andrews et al. 2014a). The traditional heritage map in Fig. 4 was created by co-author Yaqalatqa7/Johnny Jones and a version appeared in Jones (2011). We thank Catherine J. Hickson (Tuya Terra Geo Corporation) and Editor Glyn Williams-Jones (Simon Fraser University) for helpful advice as to volcanology and manuscript preparation; and Robert R. Mierendorf (U.S. National Park Service, Retired) for reprints and advice as to Indigenous use of the alpine zone in the neighbouring North Cascades region. Tamara Gorin and Dianne Ganz (Western Sky Books) are thanked for bringing to Wilson's attention local history books with two traditional stories and other information discussed here. An early draft of the paper was read by Rick Budhwa (Crossroads Cultural Resource Management). We especially thank an anonymous reviewer for a searching and very insightful appraisal that led to improvement of the paper.

This paper is based upon previously published articles, books, and consultants’ reports, as cited. It focuses upon a new interpretation of that information. No new data were generated in its preparation.

Conceptualization: MCW, BA

Data curation: BA, JJ

Formal analysis: MCW, BA

Investigation: MCW, BA

Methodology: MCW, BA

Project administration: MCW, BA

Resources: MCW, BA, JJ

Supervision: MCW

Validation: MCW, BA, JJ

Visualization: MCW, BA

Writing – original draft: MCW, BA, JJ

Writing – review & editing: MCW, BA

This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.