Japanese Swordmaking

From iron sand to finished blade — the metallurgy, craft, and connoisseurship of the nihontō.

Japanese Swordmaking / Geometry and Construction
Topics · 07

Geometry and Construction

A Japanese blade is a composite: hard, folded skin steel wrapped around a soft core, in arrangements from a simple jacket (kobuse, today's standard) to three- and four-part assemblies. The welded billet is drawn out into a bar with the blade's rough proportions, then forged into its final cross-section — ridge line, bevels, spine — and its silhouette: length, curvature, and point size. That silhouette, the sugata, is the single most historically loaded property of the blade, because it tracked how wars were fought: deep-curved cavalry tachi in the Kamakura period, oversized blades and giant points in the fourteenth century, shorter belt-worn katana when combat moved to foot, shallow refined curves in the peaceful Edo era. Appraisers read shape first; it dates a blade before any other evidence is considered.

Prerequisites: Forging and Folding — kawagane and shingane are the ingredients here. Feeds problems: hard enough to cut, tough enough to survive, knowing what you’re looking at

Practitioner

Start with the problem from Big Problems: hard and tough is a contradiction in one steel. The construction step is the structural half of the answer. The smith has two prepared steels — hard, well-folded kawagane and tough, low-carbon shingane — and welds them into a single billet whose cross-section puts each where its virtue pays:

Four lamination schemes in cross-section: maru, a single steel throughout; kobuse, a hard-steel jacket folded over a soft core, today’s standard; honsanmai, side plates and a separate edge piece around the core; and shihōzume, four different pieces boxing the core in.

The menu, roughly in order of effort: maru — one steel throughout, no core; common in daggers (thin blades have no room for a core) and in cheaper blades. Kobuse — a U of kawagane wrapped over a shingane core, so the skin and edge are one piece of hard steel; the workhorse construction of modern smiths and of history’s mass-production eras. Honsanmai — side plates plus a dedicated edge piece; more control, more welds. Shihōzume and rarer schemes box the core in with up to four different steels. More pieces buy finer control over where hardness lives — and multiply the welds, each a potential flaw. Legend credits the great Sōshū smiths with seven-piece constructions; the balance of modern evidence treats that as posthumous mythology.

With the billet welded, the smith draws it out (sunobe): hammers it into a plain bar with the finished blade’s rough length, taper, and distribution of mass, with the point cut at an angle at the end. Everything is budgeted here — the bar contains the future blade’s proportions the way a sculptor’s armature contains the pose. Then hizukuri, shaping: working section by section, the smith forges in the cross-section — the shinogi ridge line, the flat spine, the long bevels down to what will become the edge. This is freehand hammer work to tight tolerances; files and scrapers refine it, but the geometry is made at the anvil, and it’s forged very close to straight. The curve comes later, from the quench — hold that thought for the next topic.

Learn four geometry words and blades stop looking alike. Zukuri — the construction style of the cross-section: shinogi-zukuri (ridged — nearly all long swords) versus hira-zukuri (flat-sided, no ridge — most daggers). Mune — the spine’s own shape (peaked, flat, rounded; a minor school tell). Kissaki — the point, with its own facets and a dividing line (yokote) separating it from the body; its proportions run from the stubby points of early blades through medium to the dramatic elongated points of the fourteenth century. Niku — “meat”: the convexity of the bevel surfaces. More niku makes a durable, armor-tolerant edge; less makes a wicked slicer that chips easier. When people say a polisher “preserved the blade’s health,” niku is much of what they mean.

Now the payoff idea: sugata is a date stamp. Fighting style dictated form, and form changed century by century — which is why appraisal starts with the silhouette:

Era What war wanted What blades look like
Heian–early Kamakura (~950–1250) Mounted archers’ sidearm; slashing from horseback Slender tachi, deep curve concentrated near the tang, small point
Late Kamakura (~1250–1330) Heavier armor after the Mongol invasions Wider, more robust blades; stout or medium points
Nanbokuchō (~1330–1390) Field warfare escalation; intimidation Oversized blades (some later cut down), shallow curve, huge elongated points
Muromachi (~1390–1570) Massed infantry; sword drawn from the belt Shorter uchigatana/katana, curve shifted toward the tip for the draw
Edo / shintō (~1600–1780) Peace: dueling, status, aesthetics Standardized katana length, shallow refined curve, textbook proportions
Shinshintō–modern (1780–) Revivalism Deliberate re-creations of older sugata — dating gets subtler

The table is a caricature — every era has exceptions — but it’s the caricature every appraiser carries, and it works. A blade’s shape narrows its date to within a century or so before grain or hamon enter the conversation.

Expert pointers

Where practitioners argue: construction versus performance — controlled comparisons of kobuse against well-heat-treated monosteel are scarce, and honest smiths concede modern monosteel can match or beat composite blades mechanically; the composite’s case today rests on tradition, repairability, and the aesthetics of exposed skin steel. What historical blades are actually made of — X-ray, neutron, and metallographic studies of damaged blades keep finding simpler and more varied construction than the school textbooks claim, including famous-school blades that are essentially maru. Utsuri, the misty shadow above the hamon on classic Bizen blades, sits half in this topic and half in the next: partly a steel-and-forging phenomenon, partly heat-treatment, mechanism still genuinely contested.

Misconceptions

  • “The curve is forged into the blade.” The blade leaves the shaping stage close to straight; most of the final curvature appears during the quench, from differential transformation. The smith plans the curve, but water executes it.
  • “More lamination pieces means a better blade.” Each extra piece adds welds, and welds are where flaws live. Kobuse has dominated for centuries because it’s the least construction that fully solves the problem — a lesson in engineering parsimony, not a compromise.
  • “A sharper edge angle is always deadlier.” Edge geometry trades durability against slicing aggression (that’s niku). Battlefield blades carried meat on the edge for the same reason axes do; razor-thin geometry belongs to peacetime cutting practice — and to blades that chip.

Check yourself

  1. Why does kobuse construction put the fold of the hard-steel U at the edge rather than seaming two plates there? Think about where welds fail.
  2. A dagger and a katana from the same smith: why is the dagger likely maru while the katana is kobuse? Two reasons — one geometric, one mechanical.
  3. You’re handed a blade with a shallow curve biased toward the tip, 68 cm long, medium point. Which era does the silhouette suggest, and what was that shape for?
  4. Where did the fourteenth century’s taste for enormous points come from, and why did later generations cut many of those blades down? What does a cut-down blade lose forensically?

Apply it

Do sugata dating from scratch (~45 minutes): in an online museum collection with dated attributions (e-Museum, ColBase, or the Met), pull side-on photos of six to eight blades spanning Kamakura through Edo. For each, before reading the label, measure with a ruler on screen: length-to-curvature ratio, where along the blade the curve is deepest, and point length relative to blade width. Plot your measurements on a timeline, then check the labels. You will get most eras right on shape alone — and you’ll have generated a dataset that can grow into a capstone prototype (a sugata-dating chart or quiz) in Be Creative.