Chapter 2: The First Toolkit: Stone, Bone, Antler, and Wood
Era span: ~3.3 Ma (Lomekwi) – Bronze Age overlap · Difficulty: mid
Requires: Ch 1
Unlocks: Ch 3, Ch 4, Ch 6, Ch 13
Tools are the first self-replicating technology: you need a tool to make better tools. The goal of this chapter is not museum-quality knapping but a working reduction sequence — a repeatable path from raw rock to a maintained edge, plus the organic toolkit that stone alone cannot provide.
2.1 Stone Selection
Safety warning: knapping throws razor flakes and silica dust — eyes cut before pain registers, dust scars lungs over seasons, and heat-treating stone spalls explosively when rushed. Knap seated with eyes covered, strike away from hands and bystanders over a hide sheet, wet-grind or ventilate dusty work, and heat chert buried under sand with slow ramps only.
- Conchoidal fracturers (flint, chert, obsidian, chalcedony) break with sharp, predictable edges. Test by striking a fresh face: clean ripples = good; dull granular break = poor.
- Heat treatment improves marginal cherts: slow-bake buried under sand at ~250–400 °C for hours; color deepens, luster rises, flaking becomes easier. This is humanity's first materials-science process — kiln-adjacent thinking from day one (Ch 5).
- Coarse stones (granite, basalt, sandstone) serve as hammerstones, anvils, grinding slabs — durability over sharpness.
| Stone | Best use | Knaps? | Notes |
|---|---|---|---|
| Flint / chert | Blades, scrapers, arrowheads | Excellent | Nodules in chalk/limestone; heat-treatable |
| Obsidian | Surgical-sharp flakes, razors | Superb but brittle | Edge < 1 µm but chips on bone; trade good (Ch 9) |
| Quartzite | Heavy choppers, hammerstones | Poor — grainy | Use where nothing better exists |
| Basalt / greenstone | Ground axes, adzes, querns | No — grind it | Peck then grind; the tree-felling edge |
| Sandstone | Grinding slabs, polishers | No | Wet-grind to double speed, control dust |
| Slate / shale | Blanks, whetstones | Splits flat | Sharpening station stock |
Prospecting rule: look where geology concentrates silica — chalk downs, river gravels below limestone bluffs, volcanic glass flows. A day spent sourcing good nodules repays ten days of fighting bad rock.
2.2 The Reduction Sequence
- Test and select a nodule free of internal flaws (ring it with another stone; dead thud = cracked).
- Percussion flaking, hard hammer: strike platform edges whose exterior platform angle is below 90° (typically ~60–85°) with a round hammerstone to detach large flakes and set the core's shape.
- Soft hammer: antler billets remove thinning flakes with longer, shallower scars — this is the step that turns a chunk into a bifacial handaxe or preform.
- Pressure flaking: push small flakes off with an antler-tine or copper-tipped pressure flaker for final edge regularity. Pressure-flaked notches make haftable stems.
- Maintain: retouch edges in the field rather than making new tools (§2.7).
Blade shortcut: prepared cores (Levallois-style or prismatic blade cores) yield many long parallel flakes from one core — mass production, Paleolithic edition. It is a variant of stages 2–3, not a separate stage, and it pays only once a knapper can read a core reliably.
Key threshold: an edge angle of 20–40° cuts efficiently; steeper scrapers (45–70°) shave hides and wood. Match geometry to task instead of chasing one universal form.
Dead end avoided: obsessing over elegant symmetrical forms. A field-expedient flake with a good edge outperforms a beautiful biface made at triple the cost. Optimize calories-per-edge-centimeter.
2.3 Ground Stone and Abrasives
Pecking (pounding with a harder stone) followed by grinding against sandstone shapes durable axes, adzes, and celts in tough rocks like greenstone or basalt — superior to any knapped edge for felling trees. Querns and manos (grinding slabs) process grain at scale and become civilization-critical in Ch 7; flour digestion is a genuine calorie multiplier.
Sand, then later emery/corundum, are the abrasive feedstock; wet grinding doubles speed and controls dust.
Work rates (experimental baselines): pecking a basalt axe blank takes ~4–8 hours; grinding to a working edge another ~4–6. A polished axe then fells a 30 cm oak in under an hour — versus half a day by burn-and-scrape. This is why ground stone, not sharper flint, unlocks Ch 6 timber architecture: durability per swing matters more than sharpness per millimeter.
2.4 Bone, Antler, and Ivory
- Antler is springy and impact-resistant: billets for knapping, wedges for splitting wood, pressure flakers, harpoon points.
- Bone splits along its length: score-and-snap ("groove-and-split") produces blanks for awls, needles (drilled eyes with flint micro-points), fishhooks, arrowheads. Boiling in water skims grease and sanitizes; simmering scraps renders glue stock (see §2.6).
- Teeth (boar tusks, shark teeth) mount as serrated blades.
Groove-and-split in 5 steps: (1) soak bone 24 h; (2) scribe two parallel grooves with a burin along the shaft; (3) deepen to ~1/3 wall thickness; (4) wedge-lift the blank with an antler tine; (5) abrade to section on sandstone. One deer metapodial yields 2–4 awl blanks; one antler yields a billet plus a dozen tine flakers.
2.5 Wood Technology Without Metal
- Splitting: drive antler or hardwood wedges along the grain; green wood splits predictably, seasoned wood does not.
- Shaping: burn-and-scrape — char the surface in coals, scrape with a flint flake or mussel shell, repeat. This hollows canoes, bowls, and dugouts with surprising speed.
- Joining: lashings (fiber, sinew, rawhide — Ch 3) and mortise joints cut with flint chisels; pegs of dense hardwood.
GREEN LOG → wedge-split → adze to section → burn-and-scrape hollow
│ │ │ │
└─ split along grain ┌─ char 3–5 mm ─┐ repeat until
(never across) │ scrape clean │ wall ≈ 2–3 cm
└────────────────┘
2.6 Composite Tools and Adhesives
Hafting multiplies capability: a 30 cm handle turns arm motion into levered force and protects the user.
- Pitch adhesives: pine resin thickened with charcoal powder (reduces brittleness); applied hot, kneaded like taffy, hardened on cooling.
- Bitumen, where surface seeps exist, is a ready-made adhesive and waterproofing (it later seals boat seams for the river trade of Ch 9 and Ch 24).
- Animal glue: simmer hide/bone collagen for hours; strong, reversible with heat.
- Sinew wrapping, applied wet, shrinks as it dries into a drum-tight grip that mechanically locks the joint; pitch fills gaps beneath.
Key threshold: a hafted ground axe fells a sapling in minutes and a 30 cm trunk in under an hour (§2.3), versus half a day for burning or gouging — timber architecture, dugout boats, and palisades all date their feasibility to reliable hafting.
| Adhesive | Mix | Working temp | Waterproof? | Best for |
|---|---|---|---|---|
| Pine pitch | resin + ~20 % charcoal dust + beeswax if available | hot-applied ~120 °C | Yes | Hafting points, sealing seams |
| Bitumen | neat or sand-tempered | warm to hot | Excellent | Canoe seams, basket waterproofing |
| Hide glue | simmered collagen; dry glue ~1:2 with water by weight | warm pot ~60 °C | No — reverses | Furniture-grade wood joints |
| Sinew + pitch | sinew wrap over pitch bed | room temp wrap | Partial | Shock joints (axes, harpoons) |
2.7 Maintenance Doctrine
Edges are consumables. Carry a small hammerstone and pressure flaker; retouch during rest breaks. Standardize a few tool types rather than carrying dozens of specials. The toolkit that wins is the one whose repair was designed alongside its manufacture — a principle that scales unchanged up to jet engines (Ch 33).
Field kit (per 4-person team): 2 hammerstones, 1 antler billet, 2 tine flakers, 1 sandstone abrader, 3 preform blanks, 1 pitch stick, spare cord. Weight under 3 kg; rebuilds any edge in the kit in minutes.
2.8 The Deep Record
The lithic sequence is the longest dataset technology has:
| Industry | Approximate age | Signature |
|---|---|---|
| Lomekwian | ~3.3 Ma (Lomekwi 3, Kenya) | large crude flakes and anvil-struck cores; maker unknown, interpretation debated |
| Oldowan | ~2.6 Ma (Gona, Ethiopia) | sharp flakes struck from cobbles |
| Acheulean bifaces | ~1.76 Ma (Kokiselei, Kenya) | standardized symmetric handaxes — planned form |
| Prepared-core (Levallois) | by ~300 ka | core shaped to detach predetermined flakes — manufacturing planning |
| Wooden throwing spears | ~200 ka (Schöningen, Germany; revised from ~300 ka in 2025) | about ten preserved spears, mostly spruce — sophisticated green-wood working, now attributed to Neanderthals |
| Blade + microlith industries | ~45 ka onward | mass-production of standardized blades; composite microlith armatures |
Obsidian sourcing (trace-element fingerprinting) shows stone transported hundreds of kilometers from its source (Melos to mainland Greece; Lipari across the Tyrrhenian; Yellowstone obsidian across the Plains) — the oldest documented trade networks, preceding any other recordable exchange (Ch 9).
2.9 How We Know: Experimental Archaeology
Every performance claim in this chapter rests on three evidentiary methods worth naming:
- Replication experiments — knappers reproduce tools under controlled conditions and measure edge durability, cutting efficiency, and failure modes.
- Use-wear analysis — microscopic polish and scarring patterns identify which material a particular ancient edge actually cut (hide polish looks different from wood polish).
- Residue analysis — blood proteins, starch grains, and pitch traces survive on ancient tools and bind them to specific tasks.
Where claims rest on none of these, treat them as plausible reconstruction, not established fact — this discipline separates lithics from just-so storytelling.
2.10 Safety and Teaching Order
Knapping showers razor debitage: always knap seated with a hide pad over the legs, eyes protected (even a bark visor beats nothing), and novices upwind and 2 m back. Teach in this order — splitting → scraping → hard-hammer flakes → hafting → soft-hammer thinning → pressure flaking — because each stage's waste feeds the next stage's practice stock, and a student who can haft a crude flake is already useful before they can thin a biface.