How Tectonic Plates Move

For ages 11-12, teach plate tectonics as a causal heat engine, not a memorised list of boundary types. The central chain is heat from Earth's interior -> convection in SOLID, slowly-flowing mantle rock -> plate motion (driven mainly by slab-pull and ridge-push) -> boundary landforms and hazards. Confront the dominant misconception that plates float on liquid/molten mantle. Learners should drive a convection cross-section, isolate the forces that move a plate, connect the slow creep (cm/yr) to deep-time outcomes like the Wilson cycle, and layer the evidence (fossils/coastlines, seafloor age stripes, earthquake depths, seismic S-wave transmission). A manipulable 2D cross-section is the recommended core; a 3D whole-Earth view, a reference anatomy image, an expert conversation, and a consolidation check are complementary with stated limits.

• 'Plates float on a sea of molten lava / liquid mantle.' — The mantle is SOLID rock that flows plastically over millions of years; plates move on the weak asthenosphere, not a liquid ocean.
• 'Continents drift through the oceans' (old Wegener framing).' — Whole plates (crust + upper mantle, carrying both continents and ocean floor) move together; continents are passengers, not ploughs.
• 'Convection is a conveyor belt that simply drags plates along.' — Ridge-push and especially slab-pull at the plate edges dominate; simple basal drag is a weaker, incomplete picture.
• 'Boundaries are where plates start moving.' — Boundaries are where neighbouring plates INTERACT; the whole plate is in motion, and the boundary is just the seam.
• 'The mantle is molten because volcanoes erupt lava.' — Melt is localised (decompression at ridges, flux melting above subducting slabs); the bulk mantle stays solid.
• 'Plate motion is fast / you could feel plates moving.' — It is centimetres per year (fingernail growth); the drama is in the sudden slip at faults, not the steady creep.

• Mantle-convection cross-section: heat source below, cells of solid rock rising at ridges and sinking at trenches, rigid plates riding on top — the canonical driving-engine diagram.
• Ridge-push / slab-pull force diagram on a single plate: uphill slide off the ridge + downhill drag of the subducting slab, arrows summing to net motion.
• Three-boundary panel: divergent (ridge, new crust, shallow quakes), convergent (subduction trench + arc volcanoes / continental collision + mountains, deep quakes), transform (sliding, big shallow quakes, no volcano).
• Seafloor age-stripe map: symmetric magnetic stripes youngest at the ridge, oldest at trenches — the conveyor evidence.
• Wilson-cycle timeline: ocean opens at a rift → widens → subduction begins → ocean closes → continents collide → supercontinent → rifts again.

• Anchor everything to the energy source: Earth's hot interior drives convection which drives the plates — build the causal chain heat→flow→plate motion→boundary landforms, so boundary types are consequences, not a memorised trio.
• Confront the 'mantle is molten' error head-on: model the mantle as SOLID rock that creeps like extremely stiff putty/silly-putty over geological time; only small pockets melt.
• Give ridge-push and slab-pull explicit airtime rather than a single 'conveyor belt' image; note that slab-pull (dense cold slab sinking) is now considered the main driver.
• Use analogue models with named limits (a pan of simmering liquid, a lava lamp) but immediately flag where each breaks down (real mantle is solid, not boiling water) so the analogy doesn't seed the molten-mantle misconception.
• Connect scale explicitly: plates move at fingernail-growth speed (cm/yr), so the Wilson cycle plays out over hundreds of millions of years — tie the mechanism to deep time.
• Layer the evidence: matching coastlines/fossils/rock bands (drift) plus seafloor-spreading age stripes plus earthquake depth patterns at subduction zones together make the moving-plate model compelling.