1 running head: wet tectonics 2 3 10 11 12 wet tectonics ... · 24 amongst geoscientists presumes...

Wet Tectonics—Grimm 9/2004 1

Running head: Wet Tectonics 1 2 3

4 5 6 7 8 9 10 11

Wet Tectonics: A New Global Synthesis 12 13

Draft: November16, 2004 — 7000 words — 49,055 characters [including spaces] 14 Manuscript for Geology 15

16 17

Dr. Kurt A. Grimm 18 Earth & Ocean Sciences— University of British Columbia 19

[email protected] — 604-822-9258 or 604-885-0568 20 fax: 604-822-608821


Abstract 22 Introductory geology textbooks, many advanced texts and most common dialogue 23 amongst geoscientists presumes that plate tectonics is a “solid-Earth” 24 phenomenon, with fluids playing an important role in most geodynamic 25 processes. Evidence summarized below suggests that Earth’s tectonic system 26 itself is the dynamic result of water interacting with the silicate Earth. 27 The synthesis/concept outlined herein redefines Earth’s plate tectonic 28 system [the tectosphere; an open system extending from the base of the 29 asthenosphere to the top of the crust, evolving as a dynamic pattern of 30 organization that creates, orders and perpetuates itself], and elucidates 31 its specific and fundamental genetic linkages with liquid water oceans. 32 Fully interdependent and mutually-causational dynamics for the coupled 33 hydrosphere and tectosphere constitute a collectively autocatalytic system, 34 a concept of authentic circular causality that is distinct from other 35 concepts based upon cause-to-effect explanatory chains. The “wet tectonics” 36 model elucidated herein is simple, internally consistent, testable and 37 constitutes a new planetary synthesis with fundamental implications for 38 geological, geophysical, Earth system and planetary sciences, as well as 39 geoscience education. 40 41 Introduction 42 Plate tectonic theory is the foundation of modern geological and geophysical 43 sciences. In this model, radioactive decay throughout the mantle 44 [supplemented by smaller but significant thermal contributions via relict 45 heat of accretion] and heat transfer across the core /mantle boundary propels 46 heat and matter transport via convection of Earth’s mantle, through whole-47 mantle convection or a stratigraphy of smaller convective systems. Some 48 uncertain combination of processes are credited for the movement of Earth’s 49 tectonic plates upon a low velocity — and presumably low viscosity — 50 asthenosphere. In many convergent margin settings, subduction processes are 51 accompanied by dehydration reactions in the downgoing slab; this process 52 liberates water that facilitates partial melting, arc magmatism and the 53 subsequent generation of new continental crust (Press and Siever, 2001). 54 55 The critical role of water is widely recognized in various geodynamic 56 processes, including the following: magmatic systems, particularly partial 57 melting in the mantle wedge (Ulmer, 2001) and asthenosphere (Karato, 2003; 58 Hirth and Kohlstedt, 1996); metamorphism and metasomatic alteration (Hacker 59


et al, 2003a,b); crustal deformation processes [e.g. thrust faulting (Hubbert 60 and Rubey, 1959), subduction zone seismicity (Peacock, 2001), the kinematics 61 of accretionary wedges (Saffer and Bekins, 2002) and hydrothermal phenomena 62 (Stein and Stein, 1994; Fisher et al.,20o3a,b). The role of water in mantle 63 dynamics has received particular attention in recent years and will be 64 surveyed herein. 65 66 Some recent studies suggest linkages amongst climatic and tectonic phenomena, 67 including problematic linkages between Late Cenozoic mountain building and 68 climate change [(Molnar and England, 1990; Raymo et al., 1988); i.e. Did 69 accelerated orogenic rates drive Pleistocene glaciations or did glacial 70 climate create the false impression of accelerated uplift rates?]. Lamb and 71 Davis (2003) suggest that Cenozoic uplift of the high-relief Andes orogen may 72 be attributable to strong coupling across a thin veneer of subducting 73 sedimentary “lubricant”, that is in turn a consequence of cool and arid 74 coastal climates that produced low weathering rates and low rates of sediment 75 delivery to the Andean forearc. These and other similar studies lend 76 penetrating insight into the many interconnected aspects of plate tectonic 77 phenomena with water, however they shed little light upon the fundamental 78 origin and operation of the global plate tectonic phenomenon itself. 79 80 With respect to the overall operation of Earth’s tectonic system, most 81 geoscientists apparently regard water role as a peripheral or even central 82 modifier of global processes that are largely rooted in mantle convection and 83 the dynamics of interacting mantle and lithosphere (Bercovici, 2003; c.f. 84 Anderson, 2001; Hamilton, 2003). Some workers have noted very generally that 85 water may be a prerequisite for plate tectonics, (Bercovici, 1998; Regenauer-86 Lieb and Kohl, 2003; Taylor, 1999) however, a comprehensive model has not 87 been presented. I address this gap by outlining a synthesis suggesting that 88 the dynamic and persistent hydrospheric and plate tectonic systems on planet 89 Earth are fully interdependent and co-evolving phenomena. In short, we shall 90 conclude that the plate tectonic phenomenon itself is the dynamic result of, 91 participant in and perhaps driver of a circulating upper mantle interacting 92 with a liquid water envelope on planet Earth. 93 94 The paper proceeds by introducing a new heuristic [clarifying 95 characterization] for planet Earth and the plate tectonic system, introduces 96 the concepts of circular causality and autocatalysis as characteristic of 97


many dynamic systems, and then focuses on the “wet tectonics” model. In 98 particular, discussion of seafloor spreading, plate motion, subduction 99 processes and asthenosphere dynamics focuses upon the genetic coupling 100 between the plate tectonic phenomenon and liquid water oceans that permits 101 and perpetuates them both as persistent and dynamic systems. The paper closes 102 with a summary and discussion that includes a list of specific directives for 103 testing this unconventional yet compelling hypothesis. 104 105 Introducing the Tectosphere 106 The Earth system encompasses two major realms distinguished by their primary 107 energy sources (Figure 1). The ectosphere consists of the atmosphere, 108 hydrosphere and biosphere, whose dynamics are largely dependent upon solar 109 fusion and subsequent electromagnetic flux [“sunshine”] to Earth’s surface. 110 The core, mantle and tectosphere collectively comprise the endosphere, 111 energized largely by fission of radiogenic isotopes in Earth’s interior. 112 113 We define herein the tectosphere as the realm of global plate tectonics, a 114 global-scale system of plate formation, plate motion, and recycling. The 115 lower boundary of the tectosphere is defined at the base of the 116 asthenosphere, the layer in the upper mantle distinct for its low seismic 117 velocity, high seismic wave attenuation, and high electrical conductivity 118 (Karato, 2003). The upper limit of the tectosphere is defined at the top of 119 the crust. Seismically, tectosphere contains the Moho and G-discontinuities 120 and is floored by a return to faster wave velocities that characterize the 121 mesosphere. Designations of continental tectosphere and oceanic tectosphere 122 are determined by the composition and origins of the crustal component. [Our 123 usage of the term tectosphere differs entirely from the usage defined in 124 Jordan (1975), who used it to highlight unique characteristics of the 125 subcontinental lithosphere]. 126 127 The tectosphere is defined as an interactive functional unit: its boundaries 128 are distinct from compositional delineations of mantle and crust, and from 129 mechanical classification of lithosphere and asthenosphere as rheological 130 subunits of Earth’s mantle. Beyond enhancing clarity of the concepts outlined 131 below, we find the tectosphere concept very helpful for educating students 132 and the general public about Earth’s plate tectonic system. 133 134


Prominent phenomena of the tectosphere include formation of oceanic crust and 135 lithosphere (= oceanic tectosphere) at divergent plate boundaries and the 136 consumption of hydrated oceanic tectosphere in subduction zones, where 137 mineral dehydration, partial melting, magmatism, and mountain building are 138 concentrated. We assert that any comprehensive hypothesis for plate tectonics 139 must account for the dynamics and mechanism of plate motion, a phenomenon we 140 regard as an intrinsic aspect of the tectosphere phenomenon itself. In 141 addition, we aim to demonstrate that the tectosphere is correctly conceived 142 as an open system and dynamic pattern of organization whose dynamics result 143 in self-organization and self-perpetuation. 144 145 Self-Organization and Autocatalysis 146 Articulating complexity is a prerequisite to comprehending this new planetary 147 view. Consequently, we digress briefly to survey the linked concepts of self-148 organization and autocatalysis. 149 150 A Benard cell illustrates the complex flow organization that spontaneously 151 emerges within convecting fluids systems. In this experimental system, a 152 thermal gradient is applied to a thin fluid film sandwiched in glass; a 153 complex and varied array of self-ordering flow patterns appear and evolve 154 through time. 155 156 Miso soup — a common first course in Japanese restaurants and sushi bars — 157 illustrates dynamic convection processes that illuminate the plate tectonic 158 phenomenon. In it, suspended particles of miso constitute flow tracers; 159 patient observation of fluid behavior in a soup bowl reveals self-160 organization, as described below. Diapir-like columns of hot liquid rise 161 buoyantly to release heat at the soup/air interface that is visible as 162 steam. Cooler, denser liquid sinks circumferentially along the outer 163 perimeter of each rising plume, to form a toroidal (doughnut-shaped) 164 convective pillar. The interaction of numerous convective pillars produce a 165 quasi-hexagonal array of convective subsystems in plan view. In each case of 166 this ubiquitous self-organization phenomenon, the hexagonal pattern is a 167 lowest surface energy configuration, that results from vertical convective 168 processes while itself governing the positioning of these dynamic subsystems 169 through time. The dynamic association of vertical pillars and the quasi-170 hexagonal plan view pattern is one of mutual dynamic governance, that is, 171 interdependent or circular causality (Kauffman, 1995). [Note to reader: If 172


your soup does not display the behavior sketched above, send it back! It is 173 either lacking sufficient miso [viscosity too low] or it is insufficiently 174 hot]. 175 176 Articulating complexity is challenging and essential. The “chicken and egg 177 paradox” familiar to every schoolboy appears insoluble because it is imbedded 178 in a one-dimensional worldview of linear cause-and-effect chains. The science 179 of cybernetics and the broader emerging science of complex systems embraces 180 the ubiquity of circular causality for dynamic self-organizing systems 181 (Lewin, 1992). For example, autocatalysis is a category of coupled circular 182 interactions that occur in open chemical and biochemical systems that 183 develop, maintain and perpetuate themselves (Figure 2; Kauffman, 1995). 184 Specifically, autocatalysis occurs when a reaction product is itself the 185 catalyst for the reaction that created it; autocatalytic phenomena include 186 stratospheric ozone destruction (Geiger et al., 2001), the citric acid 187 (Krebs) cycle of aerobic catabolism that provides energy to each of the 188 reader’s cells (Purves, 1992.) and perhaps the origin of Life itself 189 (Kauffman, 1995; Kauffman, 2000). More widely, autocatalysis describes the 190 linkage of two or more distinct subsystems that mutually permit and 191 participate in one another’s spontaneous formation through an interconnected 192 network of interactions. 193 194 In general, autocatalytic phenomena resemble positive (i.e. destabilizing or 195 runaway) feedback systems, with the additional limitation that accelerating 196 inter-reactions are limited by the availability of reactants and/or by 197 limited production and/or transfer rates of reactants and/or catalysts. In 198 complex adaptive systems, if sufficient reactants occur in sufficient density 199 and with sufficient energy to propel their inter-reaction, the network of 200 functional interrelationships that defines the system can move to a 201 threshold, beyond which the system can spontaneously function as a collective 202 autocatalytic whole persisting as a dynamically stable entity (Lewin, 1992). 203 The dynamics of such systems at the “edge of chaos” occupy a phase transition 204 of sorts, between the static order of crystalline solids and the disordered 205 behavior alongside statistical uniformity that characterizes chaotic gas 206 mixtures (Kauffman, 1996; Chaisson, 2001). In short, autocatalysis and more 207 specifically collectively autocatalytic systems (Kauffman, 1996)— appear to 208 be a distinct signature of self-ordering, self-governing and self-209 perpetuating systems. 210


211 A more detailed discussion of circular causality and 212 autocatalytic phenomena is outside the scope of this short 213 review. For our purposes these terms are effectively synonymous. 214 Literature on these themes and broader complexity science 215 contribute to the conceptual and perceptual underpinnings of the 216 model described below. 217 218 The Wet Tectonics Hypothesis: How it Works 219 Seafloor spreading, plate motion, subduction zones and the asthenosphere are 220 four major aspects of the tectosphere. The discussion below reviews each of 221 these subsystems, with attention to the essential role of water, and advances 222 to an integrated systems view. 223 224 Free water is distinct from bound water within different geological 225 materials, where water occurs as H2O, as well as hydrogen, oxygen and 226 hydroxyl ions and a diversity of polymorphic forms (Williams and Hemley, 227 2001). For simplicity of discussion, description of water associated with 228 rocks and minerals refers to any of these species, unless otherwise 229 specified. 230 231 Divergent plate boundaries (and their associated transform faults) on planet 232 Earth are arranged along a near-continuous circumplanetary belt, much like 233 the seams on a baseball. Sea floor spreading occurs along mid oceanic ridges. 234 Hydrothermal circulation occurring along the divergent zone constitutes a 235 convective cooling process, with profound impact upon the chemical 236 composition of seawater and the oceanic tectosphere. For example, in the 237 exhalative systems associated with fast-spreading sites, cold, oxic and 238 alkaline oceanic bottom percolates vertically downward through fractures, 239 where it quenches, cools, and chemically interacts with the oceanic crust, to 240 produce a diverse assemblage of hydrated minerals (e.g. serpentine, 241 phlogopite, amphibole) as well as the hot, anoxic, and acidic hydrothermal 242 outputs that contribute substantially to bulk seawater chemistry (Elderfeld 243 and Schultz, 1996). 244 245 Water is primarily responsible for the dramatic heat loss that accompanies 246 quenching and cooling of oceanic crust and lithosphere. Convection at 247 spreading centres extend to depths approaching 500-1000m and is the primary 248


mechanism of heat loss from the slab (Stein and Stein, 1994). Beyond ridge 249 proximal processes, the presence of young (4300yBP) overpressured fluids in 250 3.5 Myr ocean crust suggests that rock/fluid interaction extends well beyond 251 the ridge axis (Cowen, 2003). Furthermore, Fisher et al., (2003a) report 252 rapid, large volume hydrothermal fluxes between seamount outcrops emergent 253 above nearly impermeable sediment cover at distances exceeding 50 km; noting 254 the commonality of seamounts and other basement outcrops on the global 255 seafloor, they credited hydrogeological interconnectivity thru bedrock 256 outcrops for striking heat flow anomalies via advective heat transfer (i.e. 257 liquid water). These suggestions of deep and widespread advection of water in 258 the oceanic tectosphere complement the conclusions of (Hacker et al., 2003a) 259 that global ocean crust is partially hydrated (<1.3 wt. % H2O) , and the 260 uppermost mantle ranges between unhydrated to 20% serpentinized (~2.4 wt. % 261 H2O). 262 263 Seafloor spreading occurs as buoyant upper mantle rises and undergoes 264 adiabatic (decompression) melting, resulting in magma genesis at shallow 265 subcrustal depths. As partial melting of rising mantle material proceeds, 266 higher water solubility in the melt phase uptakes water from surrounding 267 source rock, and through a process termed dehydration strengthening, melt is 268 focused along narrow linear belts corresponding to the ridge axis (Hirth and 269 Kohlstedt, 1996). Magma quenching occurs in the direct presence of water 270 and/or is facilitated by steep thermal gradients resulting from convective 271 and advective heat transfer. Hot, young and low-density oceanic tectosphere 272 is isostatically buoyant; as the slab of oceanic tectosphere is quenched and 273 cooled, hydration reactions accompany cooling and densification of the plate, 274 causing vertical subsidence (Stein and Stein, 1992). 275 276 The age/depth relationship of the oceanic seafloor along traverses 277 perpendicular to the spreading axis records an exponential subsidence path, 278 to a near-equilibrium depth of about 4.5 km for crust at an age of about 90 279 Myr (Stein and Stein, 1992). Subsidence is attributable to heat loss from 280 the oceanic crust and lithosphere to the overlying ocean, resulting in 281 hydration, cooling and densification of oceanic tectosphere. Seen in cross-282 section, one may envision hot young tectosphere near the ridge crest in an 283 elevated position with respect to Earth’s gravitational centre, and the top 284 of older, colder and denser oceanic tectosphere some hundreds of kilometers 285 distant about 2 km vertically downslope. 286


287 We suggest that formation and spreading of oceanic tectosphere proceeds by 288 gravitational sliding along a topographic gradient resulting from the 289 convective and advective cooling and subsidence of the tectosphere slab 290 itself; we term this gravity-driven mechanism slab pull and envision its 291 operation wherever divergent plate boundaries and sufficient liquid water are 292 present. We suggest that gravitational potential energy via mantle-derived 293 thermal buoyancy of young tectosphere may be the fundamental driving force of 294 tectonic plate motion. In addition, slab pull requires mechanical decoupling 295 of the rigid lithosphere from the low-viscosity asthenosphere, each a 296 functional unit of the tectosphere (see below). In this sense, the 297 asthenosphere may perhaps be regarded as the slippery layer upon which rigid 298 plates glide. 299 300 In short, slab pull operates via thermal densification and subsidence of 301 oceanic tectosphere, creating topography that results in extensional forces 302 propelling plate motion. Gravitational sliding of the plate along a 303 topographic gradient created by the thermal subsidence of the plate itself is 304 credited for driving plate motion and perpetuating seafloor spreading. This 305 process resembles a simple example of autocatalytic technology given in 306 Kauffman (2000). Large pumps kept otherwise flooded mines clear for miners 307 and permitted the British industrial revolution: iron and coal fueled the 308 manufacture and operation of pumps that permitted the mining of iron and 309 coal. 310 311 Subduction Zones: Trench Pull and Planetary Recycling 312 Subduction zones host underthrusting and dehydration of oceanic 313 tectosphere; resultant free water in the mantle wedge and overlying crust 314 facilitates partial melting, magmatism, plutonism, and volcanism (Ulmer, 315 2001). The distribution of water in subduction zones dramatically 316 influences wave propagation and seismicity (Bostock et al., 2002; Peacock, 317 2001), as well as electrical resistivity (Booker, 2004). In fact, 318 modeling results suggest that initiation of subduction itself may result 319 from feedback processes associated with lubrication by water (Regenauer-320 Lieb et al., 2001). Furthermore, Peacock (2001) noted that hydration of 321 tectosphere at spreading centres may be restricted to crustal levels, 322 whereas deeper (25-50km) hydration may occur by thermal convection and/or 323 seismic (dilatancy) pumping associated with fault/fracture zones 324


associated with outer rise earthquakes directly outboard of subduction 325 zones. 326 327 The magnitude of subduction and related magmatic and volcanic processes in 328 the larger water cycle on planet Earth is uncertain. Peacock (1990) 329 estimated that 8.7x10 11 kg of H20 are subducted annually, with 1.4 x 10 11 330 kg per year being returned via arc magmatism. Williams and Hemley (2001) 331 used these results to calculate that approximately 7.4 x 10 20 kg of water 332 is subducted every billion years — a value approximately equal to half of 333 the current hydrosphere — and credited subduction as the primary mechanism 334 for regassing of the mantle. Carmichael (2002) asserted that rising plumes 335 of water-rich magmas to the overlying crust and surface volcanoes 336 constitute a volatile recycling mechanism he termed “the andesite 337 aqueduct”, concluding that these return flows were approximately in 338 balance with the mass of water globally subducted. 339 340 In simplest terms, hydrospheric water —captured via mineral hydration 341 processes at spreading centers spreading and liberated via dehydration 342 reactions during subduction — permits partial melting, magmatism and rock 343 recycling that is a prerequisite to a dynamic and persistent tectosphere. 344 Plate tectonics constitutes a recycling mechanism for water and other 345 volatiles that is required for a dynamic and persistent hydrosphere 346 347 Rates of plate motion on planet Earth follow a very simple pattern. Ocean 348 basins that lack subduction zones record slow seafloor spreading (≤ 1to 349 3cm per year, e.g. Atlantic), whereas ocean basins rimmed by subduction 350 zones record rates greater by a factor of 3-5x (e.g. Pacific). Conrad and 351 Lithgow-Bertelloni (2002) compared observed plate motions associated with 352 nine subduction zones with predicted motions computed for those same zones 353 using different driving forces, concluding that the gravitational descent 354 of attached slabs accounted for about half of the driving force on those 355 plates. Further discussion subduction-related mechanical dynamics — here 356 clustered under the term trench pull — are beyond the scope of this paper 357 (c.f. Hamilton, 2003). We conjecture that motion of Atlantic and Pacific 358 tectosphere may be fundamentally driven by slab pull processes as outlined 359 in the previous section, with accelerated Pacific rates attributable to 360 trench pull. 361 362


Asthenosphere: Earth’s “slippery layer” 363 Water bound into crystal structures plays a central role in mantle 364 stratigraphy and rheology (Stevenson, 2004; Williams and Hemley, 2001), as 365 well as the genesis of mantle-derived magmas and rocks (e.g. Bercovici and 366 Karato, 2003). The asthenosphere is a distinctive layer in Earth’s mantle 367 characterized by low seismic velocity, low viscosity, high seismic wave 368 attenuation and low electrical conductivity (Karato and Jung, 1998), as 369 well as the development of seismic anisotropies in some settings (Jung and 370 Karato, 2001). Typically, the asthenosphere is not well-developed below 371 continents, whereas it is a distinct and ubiquitous component of oceanic 372 tectosphere, typically occurring between the depths of 75-150 km (Karato, 373 2003). 374 375 Early workers suggested that intergranular partial melts were 376 responsible for the asthenosphere’s distinctive physical 377 characteristics. Water was widely credited for initiating and 378 maintaining intercrystalline partial melts — and thus low 379 viscosity — in the asthenosphere. Most geological textbooks 380 perpetuate this view, however this widespread belief is not 381 supported by more recent seismological and mineral physics 382 observations (Karato, 2003). 383 384 Many of the unique properties of the asthenosphere have been 385 attributed to solid phase deformation processes facilitated by 386 incorporation of water within the crystal lattice of mantle 387 minerals, creating point and line defects. Hirth and Kohlstedt 388 (1986) discovered that the viscosity of olivine aggregates was 389 reduced by a factor of at least 140 in the presence of water. 390 Karato & Jung (1998) reviewed mineral physics observations, 391 determining that intracrystalline water significantly reduces 392 seismic wave velocity and viscosity of mantle rocks; furthermore, 393 they concluded that partial melting actually results in an 394 increase in seismic wave velocity through the transfer of water 395 from solid phase minerals into melts, suggesting that partial 396 melting in the asthenosphere occurs only in the vicinity of mid-397 ocean ridges. Karato and Jung (2002) report that strain rates in 398 olivine crystals increase a factor by 10-100 with increasing H2O 399 content, in the absence of partial melting. The broad phenomenon 400


of reduced viscosity when water is incorporated into a solid 401 phase is termed water weakening (Karato, 2003). 402 403 In the bulk mantle, most water is generally presumed to be 404 primordial (Williams and Hemley, 2001; c.f. Hamilton, 2003)), 405 whereas sources of water in the asthenosphere remains uncertain. 406 Recent studies suggest that subduction processes may be capable 407 of delivering slab-bound hydrospheric water to asthenospheric 408 depths of 75 to 150 km and perhaps considerably deeper. For 409 example, Zhang et al. (2004) utilized high pressure mineralogical 410 experiments to demonstrate that subduction-zone earthquakes at 411 depths between 100-250 km may result from dehydration 412 embrittlement and associated precipitation of water at grain 413 boundaries. Similarly, Jung et al. (2004) utilized deformation 414 experiments to infer dehydration embrittlement of hydrous 415 minerals (principally serpentine) at depths of 50-300 km. 416 417 Bostock et al. (2002) attributed slow S-wave velocities for 418 mantle wedge rocks in the Cascadia subduction zone to 50-60% 419 serpentinization, via dehydration and upward migration of free 420 water from the subducting oceanic plate. We reason that if water 421 is liberated at a P/T threshold constituting what may effectively 422 be a point source, the landward horizontal gradient of decreased 423 water content they report may reflect lateral migration of slab-424 derived water within the mantle wedge. Hacker et al. (2003a) 425 interpreted that mantle wedges locally reach 60-80% hydration, 426 while Hacker et al. (2003b) interpreted that some intermediate-427 depth earthquakes are linked to slab mantle dehydration occurring 428 at depths perhaps as deep as 180km. Peacock (2001) credited 429 dehydration embrittlement of the serpentine mineral antigorite 430 for Pacific Rim earthquakes between 100-160 km depth. 431 432 These results converge on a working hypothesis that slab 433 dehydration in subduction-zones liberates mineral-bound 434 hydrospheric water to asthenospheric depths. As a cautionary 435 note, Booker (2004) used magnetotelluric data to infer 436 subduction-related deep (≥ 250 km) melting facilitated by water, 437 however, they favored a model by Bercovici and Karato (2003) 438


positing free water liberated via mineralogical phase change of 439 slowly upwelling mantle from greater depths. 440 441 In sum, physical properties of the asthenosphere are largely 442 attributable to bound water within crystalline solids. It is 443 uncertain what proportion of that that water is primordial 444 [derived from below] or hydrospheric [derived from above]. The 445 provenance of asthenospheric water is an intriguing topic for 446 further interdisciplinary research. 447 448 Conclusions and Discussion: 449 Figure 1 presents a simple new representation of the Earth 450 system. The tectosphere is defined herein as Earth’s plate 451 tectonic system, which is an open system, a manifestation of 452 self-organizing processes, and a dynamic pattern of organization 453 that creates and perpetuates itself. The tectosphere is defined 454 as a functional entity, that extends from the base of the 455 asthenosphere to the top of the crust, and includes the processes 456 of plate generation, plate motion, as well as plate consumption 457 and recycling. 458 459 Many workers have clearly recognized that water is essential to 460 many different aspects of Earth’s geodynamism. The wet tectonics 461 model posits that the Earthly phenomenon of plate tectonics is 462 the direct result of liquid water oceans at the interacting with 463 a dynamic silicate planet. 464 465 The coupled processes of seafloor spreading (crust and 466 lithosphere formation) and slab pull (gravity-driven plate 467 motion) are mutually dependent upon liquid water: convective and 468 advective quenching, cooling (and hydration) of oceanic crust and 469 lithosphere (herein termed oceanic tectosphere) drives 470 densification, subsidence and slab pull, further perpetuating 471 seafloor spreading. Mass balance requires consumption of oceanic 472 tectosphere at subduction zones, where slab dehydration permits 473 partial melting, magmatism, plutonism, volatile recycling, 474 crustal recycling and accelerated spreading rates via trench 475 pull. Observations by seismologists and experimental 476


mineralogists are widely agreed that that low viscosity of the 477 asthenosphere requires intracrystalline water to permit solid-478 state dislocation creep; we have presented results of numerous 479 studies suggesting that asthenospheric water may be sourced from 480 hydrosphere via slab dehydration in subduction zones (cf. 481 Bercovici and Karato, 2003). 482 483 The maverick hypothesis presented herein posits that the origin, 484 evolution and continuity of dynamic and persistent tectosphere 485 and hydrosphere systems on planet Earth are linked in a circular 486 loop of mutual causality. This view differs fundamentally from 487 the view of most geoscientists — as explicitly described in basic 488 textbooks and as described or implicit in many advanced texts and 489 research articles — that plate tectonics is largely a “solid-490 Earth” phenomenon (e.g. Bercovici, 2003), and that fluids play an 491 important modifying role in most geodynamic processes. Many 492 geoscientists have accurately recognized the central role of 493 water in many aspects of plate tectonics. This model is unique in 494 its coupling of liquid water with the structure and operation of 495 the plate tectonic phenomenon (tectosphere) itself. 496 497 Simply stated, we postulate herein that the tectosphere and 498 hydrosphere are fully interdependent and mutually-causational 499 (autocatalytic) systems. In autocatalytic chemical systems, the 500 product molecules catalyze and perpetuate the reactions by which 501 they themselves are formed. We envision similar genetic linkages 502 for liquid water oceans and the global plate tectonic system 503 (Figure 3). In short, plate tectonics is autocatalytic and all 504 wet. 505 506 The autocatalytic aspect superficially resembles the water-free 507 model in Bercovici (1998) who envisioned plate tectonics as an 508 outcome of nonlinear rheological behavior resulting from the 509 feedbacks between viscous heating and temperature-dependent 510 viscosity in the mantle. We concur in many respects with Anderson 511 (2001) who regarded plate tectonic plates as an open, far-from-512 equilibrium dissipative and self-organizing system, taking matter 513 and energy from the mantle and converting it to mechanical forces 514


that drive plate motion; this alternative to the mainstream 515 “bottom-up” conception of lithospheric dynamics as a surface 516 manifestation of self-organized mantle convection processes (e.g. 517 Davies, 1999; Press and Siever, 2001; Bercovici, 2003), suggests 518 that plate tectonics is a “top-down” process, a self-organizing 519 dissipative system acting to facilitate and organize mantle 520 convection. We concur with most aspects of his dynamic systems 521 view, yet we conclude that a coupled (i.e. top-down and bottom 522 up) systems perspective better represents the tectosphere’s 523 genetic relationship with itself and perhaps portions of the sub-524 tectosphere upper mantle. 525 526 While peripheral to the autocatalytic model herein, we suggest 527 that the steady-state tectosphere may be effectively closed 528 materially, deriving heat energy from the mantle, but minimal 529 inputs of “primitive” matter from subasthenospheric depths (c.f. 530 Hamilton 2003). Larger subasthenospheric material fluxes such as 531 superplume episodes posited during Earth history (Anderson, 1994; 532 Coffin and Eldholm, 1994) and substantial changes in the 533 mechanical coupling between mantle and tectosphere (true polar 534 wander of Kirschvink et al., 1997) may be more accurately 535 understood as perturbations to steady-state tectosphere dynamics. 536 537 Hamilton (2003) posits that plate formation occurs at spreading 538 ridges when near-solidus asthenosphere melts adiabatically along 539 gaps between divergent plates (Note: more autocatalytic 540 coupling!). In his model, rifting of the oceanic plate permits 541 asthenospheric upwelling and melting that perpetuates the oceanic 542 plate; top-down cooling of asthenospheric material produces the 543 lithospheric plate. Hamilton argues that global plate circulation 544 results from subduction-related processes that are in turn 545 attributable to density contrasts attributable to top-down 546 cooling of oceanic mantle by seawater. The hinge rollback model 547 for global plate motion in Hamiltoin (2003) is compatable with 548 the autocatalytic model described herein. 549 550 Circular causality and emergent complexity is at the core of this 551 working hypothesis. In sum, we suggest that the tectosphere is a 552


self-organizing process and dynamic pattern of organization 553 producing and perpetuating itself at the interface of a liquid 554 water oceans and a dynamic silicate planet. 555 556 Testability and falsifiability is central to any new hypothesis. 557 Towards these ends, we list predictions of the Wet Tectonics 558 model to facilitate discussion and hypothesis testing: 559

1. Balanced global fluxes of water into and out of the 560 tectosphere (e.g Carmichael, 2002) 561

2. The asthenosphere —like other components of the tectosphere 562 — is generated by and simultaneously permits global plate 563 tectonics (c.f. Anderson, 2001 and Hamilton, 2003). 564

3. Substantial or even predominant quantities of water in the 565 asthenosphere are derived from the hydrosphere, delivered 566 to the mantle wedge via subduction processes and laterally 567 distributed through an as-yet unidentified mechanism (c.f. 568 Bostock et al., 2002). The mechanisms and rates of 569 bidirectional exchange posited between the mantle wedge and 570 adjacent asthenosphere and hydrosphere are very poorly 571 constrained and certainly warrant simple modeling studies. 572 We anticipate that direct fluxes of water into and out of 573 subduction zones may not balance. In fact, if subduction 574 zones do recharge the asthenosphere via vertical and 575 horizontal migration of volatiles, balanced tectonic water 576 fluxes may require estimation of subducted hydrospheric 577 water bound into slabs, liberation of this water in 578 subduction zones and return fluxes to the hydrosphere 579 through and adjacent to oceanic tectosphere (e.g. Fisher 580 etal., 2003a,b). 581

4. The origin of the dynamic tectosphere can be successfully 582 modeled as the result of a superficial mobile layer on a 583 silicate planet interacting with a liquid-water envelope at 584 the planetary surface (c.f. Regenauer-Lieb and Kohl, 2003). 585

5. The lithosphere and asthenosphere are mechanically 586 decoupled; that is, the asthenosphere operates mechanically 587 as a slippery layer upon which the lithosphere glides. This 588 mechanical interaction may be reflected in seismic 589


anisotropies directly above and below the top of the 590 asthenosphere (Jung and Karato, 2001). 591

6. Future kinematic analyses contrasting slow Atlantic versus 592 fast Pacific seafloor spreading rates may be attributed to 593 subduction-related, trench-pull processes that accelerate 594 background rates attributable to gravity-driven slab pull. 595 Alternatively, the tectonic mechanism of top-down cooling 596 of oceanic tectosphere and plate drive via hinge rollback 597 of subducting plates and associated mechanical coupling as 598 described by Hamilton (2003) is compatable with the 599 autocatalytic model outlined herein. Furthermore, since 600 fast spreading ridges are associated with very high 601 convective heat loss (Fisher et al., 2003b) and more rapid 602 subsidence of young tectosphere (Stein and Stein, 1994), 603 fast spreading may also perpetuate circular dynamics, with 604 hydrothermal cooling driving faster subsidence rates and 605 accelerated spreading. 606

7. Silicate planets possessing a convecting interior and a 607 sufficiently voluminous liquid hydrosphere sufficient to 608 participate in convective crustal cooling will develop a 609 global recycling mechanism constituting plate generation, 610 motion and consumption (Taylor, 1999; Regenauer-Lieb and 611 Kohl, 2003)) and record the association in their geologic 612 history. For example, Sleep (1994) attributed global 613 geomporphology on Mars to early plate tectonics, that may 614 have been extinguished when a liquid hydrosphere was lost 615 from the red planet (Grimm, in prep.). 616

8. The rapid self-organization and spontaneous 617 complexification recorded by dynamic autocatalytic systems 618 passing through a phase transition (Lewin, 1992; Kauffman, 619 1995 and 2000) suggest that the compositionally 620 differentiated Earth possessing a circulating mantle and 621 liquid water oceans may have developed plate tectonics very 622 early in her history. This suggestion is partly 623 corroborated by Wilde (2001) who discovered 4.4 Ga cores of 624 zoned zircons reflecting origins within a granitic melt, 625 that evolved via partial melting of a pre-existing 626 supracrustal material that experienced low temperature 627


interaction with a liquid hydrosphere. Mojzsis et al. 628 (2001) report similar findings for zircon cores dated 4.3 629 Ga. These studies do not confirm but are entirely 630 consistent with an early tectosphere (=plate tectonic 631 system) developing rapidly as the spontaeneous consequence 632 of liquid water oceans interacting with a dynamic silicate 633 planet. 634

635 Closure 636 In sum, we regard this autocatalytic hypothesis as a testable and 637 new fundamental theory of Earth and planetary sciences. This 638 hypothesis of circular causality whereby open systems mutually 639 interact to create and perpetuate a deeply coupled dynamic 640 pattern of organization that perpetuates itself differs 641 fundamentally from a linear worldview of “cause-to-effect” 642 relationships or linear cause-to effect chains. 643 644 The origins of complexity are probably simple (e.g. Lewin, 1992; 645 Chaisson, 1998, 2001), yet they manifest in a seemingly puzzling 646 diversity of pattern, creating a problematic intersection with 647 deeply inculturated linear and deterministic worldviews. For 648 example, exclusive contrasts of “top-down” versus “bottom-up” 649 alternatives for comprehension of the plate tectonic phenomenon 650 may be informed by decades of deliberation over irresolvable 651 “nature or nurture” and “chicken or egg” paradoxes. In fact, 652 these latter debates may lend little insight into the specific 653 phenomena under consideration; rather, they reveal the 654 fundamental ineffectiveness of linear cause-and-effect chains for 655 interpreting an open systems world of spontaneous 656 complexification and self-perpetuation. 657 658 Geosciences are unique amongst the natural sciences, for we 659 possess a tangible static record of genuinely complex phenomena 660 spanning a stupendous spatiotemporal range, alongside the 661 opportunity to closely observe active phenomena in natural, 662 experimental and model systems. Complexity is complex, however it 663 is not likely complicated, and it is probably simple. We 664 anticipate that the Earth sciences shall soon transcend linear 665


simplification, to embrace the uncertainty, certainty and 666 spontaneous novelty of an authentically complex world. 667 668 669 Acknowledgements: 670 Very special thanks to J. Christ. PJ Little drafted figures. UBC 671 students (especially EOSC 371 in 2003) and EOS colleagues (esp. 672 R. Clowes, L. Kennedy, J.K. Russell, and M. Bostock) provided 673 inspiration and/or assistance and clarification. Thanks also to 674 R. Behl, S. Karato, A. Maloof, E. Silver, and C. Stein for 675 helpful discussion. Funding from NSERC. 676


References Cited 677 Anderson, D.L., 1994, Superplumes or supercontinents?: Geology, 678

v. 22, p. 39-42. 679 —, 1998, The scales of mantle convection: Tectonophysics, v. 284, 680

p. 1-17. 681 —, 2001, Top-Down Tectonics: Science, v. 293, p. 2016-2018. 682 Bercovici, D., and Karato, S.-i., 2003, Whole-mantle convection 683

and the transition-zone water filter: Nature, v. 425, p. 684 39-44. 685

Bercovici, D., 2003, The generation of plate tectonics from 686 mantle convection: Earth and Planetary Science Letters, v. 687 205, p. 107-121. 688

689 Chaisson, E.J., 1998, The cosmic environment for the growth of 690

complexity: Biosystems, v. 46, p. 13-19. 691 Chaisson, E.J., 2001, Cosmic Evolution: The Rise of Complexity in 692

Nature., Harvard University Press. 693 Bokelmann, G.H.R., 2002, Which forces drive North America?: 694

Geology, v. 30, p. 1027-1030. 695 Booker, J.R., Favetto, A., & Pomposiello, M.C., 2004, Low 696

electrical conductivity associated with plunging of the 697 Nazca flat slab beneath Argentina: Nature, v. 429, p. 399-698 403. 699

Bostock, M.G., D.Hyndman, R., Rondenay, S., and Peacock, S.M., 700 2002, An inverted continental Moho and serpentinization of 701 the forearc mantle: Nature, v. 417, p. 536-538. 702

Carmichael, I., 2002, The Andesite Aqueduct: Contributions to 703 Mineralogy and Petrology, v. 143, p. 641-63. 704

Coffin, M.F., and Eldholm, O., 1994, Large igneous provinces: 705 Crustal structure, dimensions and external consequences: 706 Reviews of Geophysics, v. 32, p. 1-36. 707

Collins, W.J., 2003, Slab pull, mantle convection, and Pangaean 708 assembly and dispersal: Earth and Planetary Science 709 Letters, v. 205, p. 225-237. 710

Conrad, C.P., and Lithgow-Bertelloni, C., 2002, How Mantle Slabs 711 Drive Plate Tectonics: Science, v. 298, p. 207-209. 712

Cowen, J.P., Giovanni, S.J., Kenig, F., Johnson, H.P., 713 Butterfield, D., Rappe, M.S., Hutnak, M. and Lam, P., 2003, 714 Fluids from aging ocean crust that support microbial life: 715 Science 716

299, p. 120-123. 717 Cox, A., and Hart., B.R., 1986, Plate Tectonics: How It Works: 718

UK, Blackwell, 416 p. 719 Elderfeld, H., and Schultz, A., 1996, Mid-ocean ridge 720

hydrothermal fluxes and the chemical composition of the 721 ocean.: Annual Review of Earth and Planetary Sciences, v. 722 24, p. 191-224. 723

Fisher, A.T., Davis, E.E., Hutnak, M., Spiess, V., Zuhlsdorff, 724 L., Cherkaoui, A., Christiansen, L., Edwards, K., 725 Macdonald, R., Villinger, H., Motti, M.J., Wheat, C.G., and 726 Becker, K., 2003a, Hydrothermal recharge and discharge 727 across 50 km guided by seamounts on a young ridge flank: 728 Nature, v. 421. 729

Fisher, A.T., Stein, C.A., Harris, R.N., Wang, K., Silver, E.A., 730 Pfender, M., Hutnak, M., Cherkaoui, A., Bodzin, R., and 731 Villinger, H., 2003b, Abrupt thermal transition reveals 732 hydrothermal boundary and role of seamounts within the 733


Cocos Plate: Geophysical Research Letters, v. 30, p. 1550, 734 doi: 10.1029/2002GLo16766,2003. 735

736 Geiger, F.M., Pibel, C.D., and Hicks, J.M., 2001, The hydrolysis 737

of chlorine nitrate on ice is autocatalytic: Journal of 738 Physical Chemistry A, p. 4940-4945. 739

Hacker, B.R., Abers, G.A., and Peacock, S.M., 2003a, Subduction 740 factory 1. Theoretical mineralogy, densities, seismic wave 741 speeds and H2O contents: Journal of Geophysical Research, 742 v. 108; 10.1029/2001JB001127. 743

Hacker, B.R., Peacock, S.M., Abers, G.A., and Holloway, S., 744 2003b, Subduction Factory 2. Intermediate-depth earthquakes 745 in subducting slabs are linked to metamorphic dehydration 746 reactions. Journal of Geophysical Research, v. 108, 747 10.1029/2001JB001129 748

Hamilton, W.B., 2003, An Alternative Earth: GSA Today, p. 4-12. 749 Hirth, G., and Kohlstedt, D.L., 1996, Water in the oceanic upper 750

mantle: implications for rheology, melt extraction and the 751 evolution of the lithosphere: Earth and Planetary Science 752 Letters, v. 144, p. 93-108. 753

Hubbert, M.K., and Rubey, W.W., 1959, Role of fluid pressure in 754 mechanics of overthrust faulting: Bulletin of the 755 Geological Society of America, v. 70, p. 115-166. 756

Jung, H., Green, H.W., II, and Dobrzhinetskaya, L.F., 2004, 757 Intermediate-depth earthquake faulting by dehydration 758 embrittlement with negative volume change: Nature, v. 428, 759 p. 545-549. 760

Karato, S.-i., 1986, Does partial melting reduce the strenght of 761 the upper mantle?: Nature, v. 319, p. 309-310. 762

—, 2003, Dynamic Structure of the Deep Earth: An Interdisc. 763 Approach: Princeton, Princeton University Press. 764

Karato, S.-i., and Jung, H., 1998, Water, partial melting and the 765 origin of the seismic low velocity and high attenuation 766 zone in the upper mantle: Earth and Planetary Science 767 Letters, v. 157, p. 193-207. 768

Kauffman, S., 1995, At home in the universe The search for the 769 laws of self-organization and complexity: Oxford, Oxford 770 University Press, 321 p. 771

Kauffman, S.A., 2000, Investigations: New York, Oxford university 772 Press, 287 p. 773

Kirschvink, J.L., Ripperdan, R.L., and Evans, D.A., 1997, 774 Evidence for large-scale reorganization of Early Cambrian 775 continental masses by inertial interchange true polar 776 wander: Science, v. 277, p. 541-545. 777

Lamb, S., and Davis, P., 2003, Cenozoic climate change as a 778 possible cause for the rise of the Andes: Nature, v. 425, 779 p. 792-797. 780

Lewin, R., 1992, Complexity: Life at the edge of Chaos: New York, 781 Macmillan Publishing Company, 208 p. 782

Meibom, A., Anderson, D.L., Sleep, N.H., Frei, R., Chamberlain, 783 C.P., Hren, M.T., and Wooden, J.L., 2003, Are high 3He/4He 784 ratios in oceanic basalts an indicator of deep-mantle plume 785 components?: Earth and Planetary Science Letters, v. 208, 786 p. 197-204. 787

Mojzsis, S.J., Harrison, T.Mark; Pidgeon, Robert T., 2001, 788 Oxygen-isotope evidence from ancient zircons for liquid 789


water at the Earth's surface 4,300 Myr ago: Nature, v. 409, 790 p. 178-. 791

Molnar, P., and England, P., 1990, Late Cenozoic uplift of 792 mountain ranges and global climate change: chicken or egg: 793 Nature, v. 346, p. 29-34. 794

Murakami, M.H., Kei; Yurimoto, Hisayoshi; Nakashima, Satoru; 795 Takafuji, Naoto, 2002, Water in the Earth's Lower Mantle: 796 Science, v. 295, p. 1885-1887. 797

Parsons, B.a.S.J.G., 1977, An analysis of ocean floor bathymetry 798 and heat flow with age: Journal of Geophysical Research, v. 799 82, p. 803-27. 800

Peacock, S.M., 2001, Are the lower planes of double seismic zones 801 caused by serpentine dehydration in subducting upper 802 mantle?: Geology, v. 29, p. 299-302. 803

Press, F., and Siever, R., 2001, Understanding Earth: New York, 804 W.H. Freeman, 573 p. 805

Purves, W.K., Orians, G.H., and Heller, H.C., 1992., Life: The 806 Science of Biology, W.H. Freeman and Company,. 807

Raymo, M.E., Ruddiman, W.F., and Froelich, P.N., 1988, Influence 808 of late Cenozoic mountain building on ocean geochemical 809 cycles: Geology, v. 16, p. 649-653. 810

Regenauer-Lieb, K., and Kohl, T., 2003, Water solubility and 811 diffusivity in olivine: its role in planetary tectonics: 812 Mineralogical Magazine, v. 67, p. 697-715. 813

Regenauer-Lieb, K., Yuen, D.A., and Branlund, J., 2001, The 814 Initiation of Subduction: Criticality by Addition of 815 Water?: Science, v. 294, p. 578-580. 816

Saffer, D.M., and Bekins, B.A., 2002, Hydrologic controls on the 817 morphology and mechanics of accretionary wedges: Geology, 818 v. 30, p. 271-274. 819

Sleep, N.H., 1994, Martian plate tectonics: Journal of 820 Geophysical Research, v. 99, p. 5639-5655. 821

Stein, C., and Stein, S.A., 1992, A model for the global 822 variation in oceanic depth and heat flow with lithospheric 823 age: Nature, v. 359, p. 123-129. 824

Stevenson, D., 2004, Earth science - Inside history in depth: 825 Nature, v. 428, p. 476-477. 826

Taylor, S.R., 1999, On the difficulties of making Earth-like 827 planets: Meteoritics & Planetarty Science, v. 34, p. 317-828 329. 829

Ulmer, P., 2001, Partial melting in the mantle wedge—The role of 830 H2O in the genesis of mantle-derived 'arc-related' magmas: 831 Physics of the Earth and Planetary Interiors, v. 127, p. 832 215-232. 833

Wilde, S.A., Valley, J. W., Peck, W. H., Graham, C. M., 2001, 834 Evidence from detrital zircons for the existence of 835 continental crust and oceans on the earth 4.4 Gyr ago: 836 Nature, v. 409, p. 175-177. 837

Williams, Q., and Hemley, R.J., 2001, Hydrogen in the Deep Earth: 838 Annu. Rev. Earth Planet. Sci., v. 29, p. 365-418. 839

Zhang, J.F., Green, H.W., Bozhilov, K., and Jin, Z.M., 2004, 840 Faulting induced by precipitation of water at grain 841 boundaries in hot subducting oceanic crust: Nature, v. 428, 842 p. 633-636. 843

844 845


846 847

848


848 Figure Captions 849 Figure 1: The Earth System. The exosphere is energized by solar fluxes, the 850 endosphere by radioactive decay and relict heat of accretion. Arrows 851 constitute exchanges of matter and energy between different subsystems. 852 (Although peripheral to this study, solid arrows constitute exchanges that 853 are substantial fluxes for biogeochemical cycling at the rate scales at which 854 self-regulationoccurs; dashed arrows constitute weak or nil linkages during 855 steady state conditions (c.f. Grimm, 2003; Hamilton, 2003). 856 857 Figure 2: The autocatalysis concept (after Fig. 3.1 in Kauffman, 1995). In 858 autocatalytic chemical systems, the product molecules catalyze the very 859 reactions by which they themselves are formed. In this cartoon, the product 860 AB is a requisite catalyst for the formation of BA and the product BA is a 861 requisite catalyst for the formation of AB. For example, the citric acid 862 (“Krebs) cycle operates autocatalytically and is the prominent mechanism of 863 energy liberation (catabolism) in your body (Purves, 1992). 864 865 Figure 3: Circular causality on planet Earth. Plate generation 866 and plate motion resulting from a mobile upper layer on the 867 silicate Earth interacting with liquid water perpetuates the 868 self-organizing tectonic processes that propel and result in 869 plate motion, plate generation and crustal recycling; plate 870 tectonics permits a persistent and dynamic hydrosphere through 871 volatile recycling. We postulate that liquid water oceans at the 872 surface of the hot circulating mantle permitted and perpetuates 873 spontaneous organization of the tectosphere (i.e. Earth’s plate 874 tectonic system, extending from the base of the asthenosphere to 875 the top of the crust). In other words, the Tectosphere is a self-876 making, self-maintaining and self-perpetuating pattern of 877 organization, arising from the autocatalytic interaction of 878 Earth’s endosphere and liquid-dominant hydrosphere. The dynamic 879 and persistent hydrosphere is autocatalytically linked to a 880 dynamic and persistent tectosphere and vice versa. 881

1 running head: wet tectonics 2 3 10 11 12 wet tectonics ... · 24 amongst geoscientists presumes...

Documents