Terrouge E-zine Archives
Select a different issue | Select a different title
The Geology of Mossflower
By: Mitya Shostak (Guest)
Volcanic fortresses. Bottomless crevasses. Hidden islands. Dancing cliffs. Throughout the Redwall novels, the very countryside of Mossflower has served not merely as a backdrop to the action but has also played an active role in the series' many plot points - a narrative service that simple, homogenous flatlands could never have provided. The widely varied topographical features of Mossflower have imbued those woods with a character as rich as that of any dashing hero, blackhearted villain or humorous sidekick. By looking at those features in their "current" states in the course of the series, their geological backstory can be pieced together into a tale that is as tumultuous and exciting as any creature's lifetime, though it took an uncountable number of those lifetimes to unfold.
That story does not begin with Salamandastron, but that's as good a place as any to begin deciphering it, just as with so many other mysteries in the books. With a name meaning "Mountain of the Fire Lizard," and the positioning of the Badger Lords' forge room so that the flames are visible coming out the top of the mountain, it's clear that the earliest settlers of the fortress had the impression of it as being volcanic. The conical shape of the mountain and the fact that it possesses a deep central chamber makes it more obvious that Salamandastron was, at one point, an actual active volcano.
But volcanoes seldom occur as individual entities (unless they are the first product of a hotspot in the middle of a tectonic plate, and even those eventually evolve into chains as the plate moves over the hotspot, as is the case with the Hawaiian islands). Looking up the western part of Mossflower from Salamandastron, though, we see ample evidence of past volcanic activity. The caves in the northern part of the mountain range and cliffs lining the coast could have been excavated by explosions, they could be inactive volcanic fissures, or they could be dark places between fallen and eroded volcanic rocks rather than proper caves.
I would conjecture, however, that Pearls of Lutra's Ruddaring, High Rhulain's Green Isle, and Salamandastron and Marlfox's Great Inland Lake and its hidden island are also all volcanic remnants. The structure of Ruddaring, with its appearance as a solid mountain from the outside and its lowered secret chamber inside, is like that of an eroding volcano with its vent hollowed out. The otter civilization is built on a solidified core of volcanic rock, and the narrow hidden entryway is likely a smaller vent also hollowed by explosion and erosion. The Great Inland Lake may seem less inherently volcanic, but that belies its having been, if anything, far more explosive than Ruddaring or Salamandastron. There are plenty of recorded cases of volcanoes completely self-destructing: Mt. St. Helens decapitating itself in 1980, or Krakatoa blasting itself out of existence in 1887. Less renowned in pop culture are Indonesia's Mt. Toba and Oregon's Mt. Mazama, but the catastrophic demises of these two volcanoes formed Lake Toba and Crater Lake, respectively. Both of these lakes are enormous, occupying the space of the former mountain, and both contain islands that are remnants of the mountains' cores (indeed, Krakatoa also left island remnants, though it was surrounded by water to begin with). Mossflower's Great Inland Lake and its hidden island fit this model perfectly. Whatever volcano was once there was surely massive, and its paroxysmal eruption undoubtedly had widespread effects beyond the Mossflower region. There is a crater lake on the Green Isle as well; it was likely caused by the same process as the Great Inland Lake, only on a smaller scale. Red rock particles scattered by those colossal eruptions would have settled over the entire region and worked their way into sediments everywhere, contributing to the formation of the red sandstone for which Redwall is famous.
As previously mentioned, chains of volcanoes can be formed by hot spots in the middle of a moving tectonic plate, but hotspot volcanoes ooze and sputter rather than exploding spectacularly. Explosive volcanoes are caused by subduction - when an oceanic tectonic plate meets with a land plate and is forced underneath it. The pressure of the one plate under the other, plus the water seeping into the place where they meet, forces magma upward to cause chains of volcanoes along the length of the plate boundary - such as with the Andes in South America, the Cascades in the Pacific Northwest, or the entire set of Indonesian islands. The placement of all Mossflower's former volcanoes relatively close to the coastline further supports that the area was shaped by subduction.
All of this volcanism is well in the past by the time the Redwall series begins, however, and while individual volcanoes within a system may go extinct even as others remain active, the fact that the entire Mossflower region has been quiet for so long that the former volcanoes are safe to inhabit is a fairly clear indication that the subduction process has stopped. In a straightforward situation, oceanic plate after oceanic plate will subduct under a land plate; something has to occur to change the nature of the boundary. In California, the switch away from subduction occurred when the tip of the Pacific Plate, jutting between the tail ends of two smaller subducting plates, ran into the North American continent at such an angle that it started grinding alongside the North American Plate rather than sinking under it. As the smaller plates to either side subducted further, the length of this side-by-side slipping zone - called a transform boundary - increased, becoming the infamous San Andreas Fault.
I propose that something similar happened in Mossflower country. This kind of boundary can accommodate active, or even rapid, plate motion without any volcanism. There will be occasional, very large earthquakes, but on a mature fault system, they are far between; most of California's earthquakes are not on the San Andreas, which moves in sections, each of which ruptures on average of every 150 to 300 years. Regardless of whether the lifespans of Mossflower's creatures are comparable to wild animals or to humans, many generations of stories could still have unfolded there in that interval between major earthquakes; this plate boundary system can still be active without having acted up in the course of the series. Looking at the maps in all of the Redwall books - the wonderful example from the German edition in particular, which seems to have painstakingly compiled important places from all of the books into a single map - further argues for the presence of a transform fault in Mossflower. The River Moss, as well as all the other streams in the region that extend far enough west, all exhibit a point along their paths where they divert to the north/right for a short distance before continuing west toward the sea. On the German map, those points of diversion are connectable by a surprisingly straight line. Rows of diverted streams are a hallmark of transform faults, serving as a marker of the main fault trace and also as sites to study in an effort to understand past motion along that fault.
The trace of the transform fault as delineated by diverted streams and rows of mountains does not take it anywhere near the ruins of Loamhedge, with their cliff and crevasse, the region that experiences an earthquake at the end of Mattimeo. The actual placement of this area is vague compared to the rest of Mossflower's geography, but it is known to be south and east of Redwall; in the context of the rest of the geology as described here, it makes sense that it would be in line with the Great Inland Lake, only further east. Without the cliff and crevasse, it could be argued that the quakes out there, on that line perpendicular to the main fault, are still related to it; southern California's transform Garlock Fault runs perpendicular to the San Andreas, and developed to accommodate some of the pressure caused by a bend in the San Andreas where it meets the Garlock. Cliffs and crevasses, however, are not at all characteristic of transform faults, and are instead far more typical of rift zones.
A rift zone is either where two tectonic plates are pulling away from each other, or where a single plate begins to break into two pieces. The rift between the North American and Eurasian Plates is readily visible at Þingvellir, Iceland, in the form of steep rocky cliffs and plateaus riddled with deep cracks, very reminiscent of the way the scene around Loamhedge's remains is described. However, many of the fissures around Þingvellir are spectacularly volcanic, spewing walls of lava miles long, and the cliffs and fissure around Loamhedge have clearly been inactive enough that civilization (Malkariss) could be built within the crack. I propose that, at one point in the past, the solid plate around Loamhedge began to pull apart, producing that single cliff and crack. This rifting could have contributed to the potency of the volcano that became the Great Inland Lake, since it would have brought more magma closer to the surface than the subduction zone alone, and a bigger magma chamber means the potential for a bigger explosion. But at some point after that massive eruption, some force - perhaps the change in plate pressure caused when the transform fault took over from subduction - caused the rifting process to cease rather than to progress. Even failed, rifts have their fair share of seismicity. The New Madrid Seismic Zone in southern Missouri and western Tennessee is the site of a failed rift that is deeply buried under many layers of newer rock and soil, but it still produced a series of four magnitude 7 and greater earthquakes in the winter of 1811-1812. The failed rift around Loamhedge is far more exposed than the faults at New Madrid, and is a clear culprit for those "dancing cliffs" that the Stump family describes.
By figuring out the histories of individual geological features within Mossflower, the order of the larger-scale events that shaped the entire region come into better focus. The subduction started first and lasted for quite some time. The volcanoes that became Salamandastron, Ruddaring, and all the others formed and erupted, in some cases partly or fully self-destructing before growing and erupting again. At some point during this phase, the rifting began out by Loamhedge, causing many earthquakes and fissure eruptions. The rift fed into the Great Inland Lake volcano, causing it to completely destroy itself. Not long thereafter, the plate boundary began to shift from subduction zone to transform fault. The resultant changes in pressure stopped the rifting, and the change in boundary type slowly caused the volcanoes to go extinct. The transform fault continues to shape the landscape while erosion takes its toll on the volcanoes, and both it and the failed rift continue to cause sporadic earthquakes in the region.
And after all of that, woodlanders and vermin came to Mossflower, Salamandastron was converted into a fortress, Kotir was founded and destroyed, Redwall was built, and many generations of battles and quests ensued. Perhaps these events would have occurred differently if the countryside itself had a different history. Perhaps future events in the world of Mossflower will be shaped by further changes in that landscape.