Konoha-Gata

Identity

Name. Konoha-gata noko (木の葉型鋸; "leaf-shaped saw"). The Takenaka Carpentry Tools Museum history uses 木の葉型鋸 [KN1], and the 1989 report title and abstract use いわゆる『木の葉型鋸』, "the so-called konoha-gata noko" [KN2]. The consulted modern museum and research sources use this descriptive label; this review did not establish whether it was a period term. A shorter variant, 木の葉鋸 (no 型), appears in trade-history literature [KN6]. 木 ko/ki = tree; の葉 no ha = leaf; 型 gata = shape/type; 鋸 nokogiri/noko = saw. (The lineage tree and the Kataba entry use 木の葉鋸; a spelling to reconcile on a future tree refresh. Flagged, not edited.) (Terminology note: "pure pull," "pure push," "mixed-stroke," "kataba" as used here are Haft's analytic labels, not established period classifications.)

Family / category. Japanese saws (nokogiri); an early and influential medieval Japanese crosscut (横挽 yokobiki) form. Single-edged, leaf-shaped, attested from the medieval period (picture scrolls) with use apparently continuing to the mid-Edo period [KN1]. Not a rip tool by the evidence: the test favored crosscutting, and the museum history says Japan had no dedicated rip saw until around the mid-Muromachi period [KN1]. It is a plausible morphological predecessor of later curved crosscut forms such as the Anahiki, and part of the background from which later specialized saw forms emerged; documented direct descent is not established by any source cited here.

One line. The Japanese medieval leaf-shaped crosscut saw, a curved, single-edged plate tapering to a point, known from picture scrolls and a few excavated examples, and the subject of a 1989–91 museum-bulletin reconstruction series whose replicas performed better crosscutting than ripping and were used with a substantial push component as well as pull.

What the evidence supports, stated plain. (1) A leaf-shaped medieval crosscut saw with curved tooth line and pointed tip is attested in picture scrolls, apparently used to the mid-Edo period, with few excavated examples [KN1]. The Takenaka authors themselves open by saying it is "depicted prolifically in medieval picture scrolls, but the details remain unclear" [KN2]. (2) In a seven-person test of six modern SK-5 reconstructions based on a saw excavated at Kusado Sengen in 1980, testers used substantial push as well as pull action. This supports mixed-stroke use as a plausible historical interpretation, but it is not direct evidence of medieval technique [KN2]. (3) Separate research infers that pull-use became general in the medieval period and places clearer tooth-form and material evidence for pull around the fifteenth century [KN5][KN18, relayed]. The konoha-gata is relevant to that transition. The present evidence does not show that it was the single ancestor of later Japanese crosscut saws, that the family followed one convergence path, or that any stroke practice passed from it to the Maebiki-Oga. Name note: 木の葉型 is a shape name used across different objects. Kofun-period finds also called 木葉型 are tiny (about 10 cm) instruments probably for metal or antler [KN7]. Keep them distinct; whether the naming relationship to the medieval building saw is more than a shared silhouette is unresolved.

✦ How NOT to use it

Five misuse classes. Because the tool is extinct, several are reasoned from the replica test and from general hand-saw practice, said plainly.

Not-for (jobs it was shown to be poor at)

• Ripping. The replicas were better suited to crosscut, and large use of the tip was difficult, especially in ripping [KN2]. Boards in the era were split with wedges, then sawn with the imported ōga [KN1].
• Fine joinery or any finished cut. A later specialized saw's job.

Dangerous (misuse that injures the user)

Evidence status, plainly. No konoha-gata-specific operating manual, validated modern product guidance, or injury account was found. That is a research result, not a risk assessment. The best available direct test is performance research [KN2]; it is not a safety validation. It does not establish safe plate steel, heat treatment, tang construction, tooth set, handle retention, stock range or inspection limits for a new replica, and historical reconstruction specifications are not a complete safe-build or safe-use recipe. Any replica is an experimental object.

What the hazard actually is at this scale. The tested stock was 45 × 105 mm hinoki [KN2]. At that scale the hazards are those of any sharp, flexible, toothed blade driven by hand against a workpiece: contact with the teeth, a plate that buckles or binds, a handle or tang that works loose, a damaged tooth line, and stock that moves under the cut. A replica built to a thin historical gauge adds uncertainty in plate strength that no source here resolves.

Controls for any replica test (general hand-saw precautions per the review; not drawn from a konoha-gata source and not re-read in an outside safety source this pass):

• Competent supervision, and a test plan appropriate to the actual stock, not borrowed from the 1989 test.
• Secure workholding and support of the actual stock, so it cannot walk or drop at the end of the cut.
• Hands and body kept out of the blade path; the off hand held clear of the line of cut and the end of the stroke.
• Eye protection.
• Inspect the plate, tooth line, tang and handle before each use; stop if the plate buckles, binds, cracks, kinks or sheds teeth.
• Handle the blade by the handle; protect the edge in storage and transport.
• Follow the maker's limits for the steel and construction if the replica comes from a maker; none exist for a self-built one.

If bucking loaded timber (separate case, not the governing hazard). The USDA Forest Service and PCTA crosscut bucking and limbing guidance [KN14] concerns logs and limbs where compression, tension, a closing kerf, spring-back and wedging matter: anticipate which side is in tension and which in compression, watch the kerf, wedge accordingly, and expect tension wood to separate with force when cut [KN14, as quoted in the original entry]. This applies when loaded logs or limbs are actually being bucked. It should not be generalized to ordinary joinery-size stock, and it is not evidence of a konoha-gata accident pattern.

Damaging (misuse that ruins the tool or the work)

• Rust on an unsheathed plate. Japanese high-carbon saw plates corrode if stored bare; the tradition keeps them dry and oiled [KN12]. Archaeological survival patterns depend on soil and site conditions and should not be read as a single cause.
• Wrong sharpening, if the plate was laminated. A laminated hard edge is destroyed by over-grinding [KN16]. Conditional: the medieval saw's construction is not established.

(The earlier claim that a kinked plate cannot be straightened to tolerance without specialist hammer-work is deleted: no source supports it.)

Beginner-trap (mistakes that look fine until they aren't)

• Treating test behavior as technique. Replica testers used push as well as pull; that is a reason to be curious, not an operating instruction. Choose stroke direction from the replica's actual geometry and test plan.
• Sweeping the whole blade. The report says the saw was not used by sawing through in one sweep from heel to tip, and the tip was used only for fine work [KN2, as quoted in the original entry].
• Unanchored stock. Leaving the stock unsupported so it moves under the stroke is a general hand-saw trap; the bucking sources [KN14] are an analogy, labeled.

Lineage-misuse (treating it like another saw)

• Reading it as the same as the modern pull saws or as the push Maebiki-Oga. The modern finishing saws are pull saws; the maebiki is a later rip saw. The coexistence of push-and-pull behavior in one reconstruction and later push-use in the maebiki invites comparison, but no transmission between the forms is established.
• Expecting the Anahiki's coarse-bulk capability. Do not project the descendant's later tooth or heft backward.
• Confusing it with the Kofun 木葉型 micro-saw. Same silhouette name, different object [KN7].

Correct use (brief, the workbench holds the deep operation)

What it appears to have been suited to. Principally a crosscut form: the reconstruction performed better crosscutting than ripping, and its tip was especially unsuitable for large rip strokes [KN2]. It was most efficient near the blade center, and was not used for large strokes at the tip. A full heel-to-tip sweep is not what the test describes.

Stroke, in the replica test. In a seven-person test of six modern SK-5 reconstructions based on the Kusado Sengen find, testers used substantial push as well as pull action (the abstract: 引き使いだけでなく押し使いの比率も高い) [KN2]. This supports mixed-stroke use as a plausible historical interpretation, but it is not direct evidence of medieval technique. Do not instruct a modern user to push and pull a replica merely because test participants did. Stroke direction depends on the actual tooth geometry, plate stiffness, construction and the replica test plan.

Matching

Does. Crosscuts timber across the grain in medieval Japanese building carpentry [KN1][KN2]. In the reconstruction test it performed better crosscutting than ripping, was used with substantial push as well as pull action, was most efficient near the blade center, and its tip was hard to use for large strokes, especially in ripping [KN2].

Suits. Historical research, museum interpretation, experimental archaeology, and expert-led reconstruction projects. It is primarily a historical/reference and reconstruction subject, not an ordinary modern purchasing recommendation. (Material scope is thin: the direct evidence concerns hinoki of one cross-section, 45 × 105 mm; performance on hardwood, green timber, large logs or compound grain is not established.)

Kit-fit. No commercial or living-tradition source. A working shop's crosscut needs are met by modern Japanese saws, and the right one depends on the job: fine crosscut (Dozuki, Kataba), general crosscut (Kataba, Ryoba), heavy bucking (Anahiki), rip or resaw (Ryoba rip edge, Maebiki-Oga). These are not automatically interchangeable with each other or with the konoha-gata. A historical or reconstruction project is a legitimate match; see the safety section before building a replica.

Not-for (project-keyed).

• Ripping and resawing: the test found the form better suited to crosscut, with large tip use difficult especially in ripping [KN2]. Treat it as principally a crosscut form; the evidence does not prove no historical user ever ripped with it.
• Fine joinery or finished tenon-shoulder work: later specialized saws were built for that.
• Heavy log-end bucking at the Anahiki's scale: nothing cited shows the konoha-gata had the anahiki's coarse tooth or heft [KN3].

Access. None. The tool is extinct and no public accession-level object record for the Kusado Sengen saw has been located. The 1989–91 report series documents tested plate thickness, tooth angle, rake and handle dimensions, but it is performance research, not a safety validation or a build recipe (see Safety).

Revival. A test replica following the series is the only route to hands-on experience; there is no commercial source. It must be treated as an experimental object under the safety section below.

Anatomy

(The anatomy rests mainly on the 1989–91 reconstruction series [KN2][KN3][KN4], a museum research-bulletin series. Every figure from it is a result for those replicas and that test, not a specification of the medieval object. Where a point is general Japanese-saw norm, it is flagged.)

The plate. Leaf-shaped, single-edged, tooth line curved, tip pointed [KN1]. The Watanabe study reports the blade path shifting in antiquity from concave to convex [KN5] [review, relayed from the reviewer's source note: the report supports this as inference].

Plate thickness (scoped to the test). In this reconstruction test, plates at or below 1.1 mm were judged impractical because of buckling [KN2]. That places the practical threshold above 1.1 mm for those replicas; it is not a minimum to build to, not a historical dimension, and not a prescription. The result is specific to the reconstructed geometry, SK-5 steel, the 45 × 105 mm hinoki and the test method.

The tooth. The Part-2 report tested an isosceles tooth form filed in the central blade [KN3]. The review notes that the Part-3 abstract names the test tooth as a 55° isosceles 箱屋目 form, with four nageshi angles and a proposed optimum of 35–40° for that configuration, and also reports no significant difference among the four saws, although the 40° saw rated highest [KN4] [review, relayed]. "Isosceles" describes geometry only; 箱屋目 is a Japanese tooth term that should be translated and explained by a specialist, not reduced to geometry. Do not back-project the descendant anahiki's ibarame tooth onto the medieval saw.

Set, tang and handle. The reconstruction tested a handle of minor diameter 22 mm, major 28 mm, length 200 mm [KN2]. The asari set, tang seating and handle species (paulownia, cypress) are general Japanese-saw norms [KN12], not separately sourced for this saw. (Tooth, handle and set for the medieval object itself are UNKNOWN.)

Construction. Earliest Japanese saws were simple strip iron [KN1]. The replicas used SK-5 modern tool steel [KN2]. Whether the medieval saw was simple iron or laminated is not established by any source cited, and is left open.

Variations (and what it is NOT)

• Konoha-gata noko / 木の葉型鋸 (木の葉鋸): the medieval leaf-shaped crosscut saw. No named sub-forms appear in the sources [KN1][KN2].
• Not established as the Kofun 木葉型 micro-saw. Kofun finds called 木葉型 are about 10 cm and thought to be for hard materials such as metal or deer antler [KN7]. Keep the two distinct; whether the medieval name is a deliberate enlargement of the same form-name is unresolved.
• Later curved crosscut saws such as the Anahiki: resemble it in silhouette; relationship as per Lineage.

Trades (where it appears, with trade-specific notes)

• Medieval building carpentry (番匠 banshō / 木工 moku): the home trade, as a crosscut saw of medieval building work [KN8][KN1].
• General timber work: alongside the axe, adze, wedge and froe in the pre-rip-saw kit [KN1][KN7].
• Iconographic "trade," the picture scrolls: the saw's main surviving habitat [KN2][KN7][KN10].

Lineage

The konoha-gata is an early medieval Japanese crosscut form. Later curved crosscut saws, most visibly the Anahiki, resemble it in silhouette, and later specialized forms emerged from the same background. This section records resemblance and chronology. It does not assert documented descent. The rip Maebiki-Oga sits on the imported ōga rip-saw branch and is not placed as a konoha-gata descendant.

Antecedents

Before the saw: the axe and the wedge. Japan's prehistoric woodworking tool was the stone axe; iron arrived in the Yayoi period, and the iron saw appears considerably later, in the early Kofun period [KN1]. Boards were obtained by wedge-splitting until the mid-Muromachi rip saw [KN1]. The pre-saw work sequence of fell, adze (釿 chōna), spear-plane (槍鉋 yariganna) is reported in [KN7].

The Kofun strip-iron proto-saw. Per the Takenaka history, early saws were "simple plates of strip-shaped iron with plain teeth filed in" (短冊型鉄板に素歯を刻んだ簡素なもの), whose building use is "questionable"; by the late Kofun period saws gain a tang, asari (set) and nageshi [KN1]. The leaf curve is a later, medieval development.

The Nara crosscut form: the Hōryū-ji treasure saw (N-141). The oldest surviving transmitted Japanese saw, a single-edged Nara-period crosscut saw of about 648 mm with isosceles-triangle teeth (TNM N-141, Important Cultural Property 工第922号) [KN1][KN9]. It shows crosscut function was established before the medieval period; it is a flat Nara form, not a konoha-gata.

Contemporaries and later forms

• The Anahiki (鼻丸 hana-maru, "round-nose"): a later curved crosscut saw whose round nose resembles the leaf curve. The anahiki entry (Run 141) proposed the konoha-gata as a shape predecessor; a plausible morphological predecessor is as far as this entry goes. A separate Takenaka bulletin source may support a direct relationship, but it is not cited here for that claim, so it is not made.
• The later yokobiki Kataba line: Edo-period specialization diversified saw forms [KN1]. This entry does not claim the konoha-gata is the common ancestor of every named crosscut saw.
• Wedge-splitting: the board-making method of the same era, not a saw [KN1].
• The continental ōga rip saw: arrived around the mid-Muromachi period and is the foreign branch of the family [KN1][KN6]; the Maebiki-Oga belongs to it.
• Mechanized sawing (Meiji on): the eventual end of the hand tradition at scale [KN1].

History

Crosscut saws came first; the rip saw came late and foreign. Japan had crosscut saws from the Kofun period but no rip saw until about the mid-Muromachi period; the Takenaka history states that until then "Japan had no rip saw and lumber was produced by the splitting method using chisels" [KN1][KN6]. (This sentence is carried from the original entry's quotation of [KN1]; this pass re-read only the konoha-gata sentence on that page.)

A medieval form known from scrolls and few finds. The Takenaka history, re-read this pass: 「中世に入ると、鋸歯を湾曲させて先端部が尖らせた『木の葉型鋸』が用いられたことが絵巻などからわかる。この形式の鋸は江戸時代中頃まで使用されたと思われるが出土例は少ない」 ("Entering the medieval period, we know from picture scrolls and the like that the 'konoha-gata noko', with the saw teeth curved and the tip pointed, was used. This form is thought to have been used until the mid-Edo period, but excavated examples are few") [KN1]. The museum says medieval; the reconstruction authors' Kamakura attribution is theirs and is cited as such.

Pull-use in the medieval transition. The 1999 Watanabe report uses cautious language (推定される, "is inferred") for the medieval generalization of pull use [KN5]. The 1997 Takenaka report is more precise: material evidence of a push/pull distinction becomes clear around the fifteenth century, crosscut teeth change toward pull-cut ibarame around then, and pictorial confirmation of pull use appears from the seventeenth century [KN18, relayed]. These are separate evidentiary milestones, and none documents the konoha-gata's own stroke.

Edo specialization. The Takenaka history: with the rise of tool-production capacity and the division of crafts in the Edo period, saws subdivided by use, and most traditional forms seen today came into being, with production centers at Aizu and Miki [KN1]. The passage documents diversification; it does not say what became of the konoha-gata, and this entry does not claim it was displaced by the specialized forms. That reading would be inference only.

Petroski lens (interpretive, labeled). A general curved plate cannot be optimal at the fine joinery cut, the heavy bucking cut and the plunge cut at once. This is Petroski's "form follows failure" applied as a lens, not a documented causal history.

✦ Human side

The trade: the 番匠 (banshō), the medieval building carpenter. Per a general banshō history, the banshō (also 木工 moku) were the building craftsmen of medieval Japan, precursors of the daiku; they worked under state and temple building offices, independent workers multiplied by the Kamakura period, the daiku-shiki headman office crystallized between the 13th and 14th centuries and the Muromachi shogunate moved against it in 1510; they formed no guild (座) [KN8]. (Trade context only, sourced to a general encyclopedia article that does not name the konoha-gata; background, not konoha-gata evidence.)

The picture scrolls. The Takenaka authors state the saw is depicted prolifically in medieval picture scrolls and that the details remain unclear [KN2]. The Japanese Wikipedia saw article names the 石山寺縁起絵巻 as showing a Kamakura-period crosscut saw in a construction scene [KN7]. The 春日権現験記絵 (1309) is a tool-rich scroll often used for medieval carpenters' tools [KN10]; a firsthand exhibition account did not itemize a saw, so a konoha-gata identification there is plausible but not confirmed.

Why the leaf shape: one hypothesis. A modern blog suggests the arm traces a curve and curved saws follow it [KN11]. Low-confidence source; an ergonomic hypothesis, not a demonstrated design rationale. What the test supports is centre-of-blade efficiency and limited large-stroke use of the tip [KN2], without adding an unreported stroke length.

Named hands. No named maker was identified in the consulted sources. This is a search result, not a finding that none existed. Yoshikawa Kinji's 『鋸』 (1967/1976) is the canonical saw-history monograph, cited by the reconstruction report [KN2][KN13]; it was not read this pass.

Texture. A blog account of the Kasuga Gongen scroll describes tanned, muscular banshō and child apprentices carrying shavings [KN10]. Held as texture, labeled.

✦ Proof of existence

An excavated medieval saw and a museum-bulletin reconstruction series. A medieval saw was excavated at Kusado Sengen (草戸千軒町遺跡), Hiroshima Prefecture, in 1980 and is treated by the reconstruction authors as Kamakura-period [KN2][KN19, relayed]. The Hiroshima museum dates the site, not the saw, to the mid-thirteenth through early-sixteenth century [KN15, relayed]; a site-wide occupation range is not an artifact date. None of the records consulted supplies the object's layer, feature, associated finds or catalogue record. The Takenaka Carpentry Tools Museum research bulletin (竹中大工道具館研究紀要) published the series: Hoshino, Hirasawa, Watanabe and Tsuchiya (1989), Vol. 1, pp. 1–13, then a second report (1990) and a third (1991). J-STAGE classifies these as research/technical reports (研究報告書・技術報告書), confirmed for the 1989 record this pass. Open access and a DOI do not establish external peer review, and none is claimed. The report as relayed from its abstract: based on the 1980 find, six reconstructions with the same external form, dimensions, tooth shape and angles but differing plate thickness were made and tested by a seven-person panel [KN2].

Museum attestation of the form, era and use. The Takenaka history page, re-read this pass, states the form, its picture-scroll evidence and its use to the mid-Edo period [KN1]. It is excerpted from the museum's Collection Catalogue No. 1 (Saws) per the page [KN1, as carried].

Function attestation, not the leaf form. The Hōryū-ji treasure saw (N-141) and crosscut saw-marks on a Hōryū-ji bracket block attest early crosscut function, not the leaf form [KN1][KN9].

What is not in hand. No public accession-level record of the Kusado Sengen saw was located. The Hiroshima designated-objects catalogue lead [KN20, relayed] surfaced a leaf-saw-shaped item that is a wooden object, not the iron saw, and must not be substituted for the missing metal-object record. No accession number is invented here. Route to firmer: locate the Kusado Sengen object record (a 文化遺産オンライン or Hiroshima museum record) and a verbatim picture-scroll identification.

references

```
[KN1] Takenaka Carpentry Tools Museum (竹中大工道具館), 鋸の発達史 ("History of saw development").
https://www.dougukan.jp/tools/12 — kind: museum. The konoha-gata sentence RE-READ
this pass (2026-10-16); other passages carried from Run 142. Supports: medieval leaf
form, curved teeth, pointed tip, picture-scroll evidence, use apparently to the
mid-Edo period, few excavated examples, late arrival of the rip saw, Edo
diversification. Does NOT support complete direct ancestry or any maebiki link.
[KN2] Hoshino, Hirasawa, Watanabe & Tsuchiya (1989), "わが国中世のいわゆる『木の葉型鋸』について",
竹中大工道具館研究紀要 Vol. 1, pp. 1–13. DOI 10.50862/dougukan.1.0_1.
https://www.jstage.jst.go.jp/article/dougukan/1/0/1_BTCTM0101/_article/-char/ja/
— kind: museum-bulletin research/technical report (J-STAGE class 研究報告書・技術報告書,
confirmed this pass; NOT shown to be peer-reviewed). Opening sentence confirmed this
pass. Rest of abstract summarized by the fetch tool this pass and carried from Run 142
[manual, relayed]: 1980 Kusado Sengen find; six SK-5 replicas; 45×105 mm hinoki;
seven-person test; better crosscut than rip; high ratio of push as well as pull;
centre-of-blade efficiency; plates at or below 1.1 mm impractical for buckling;
handle 22×28×200 mm. Reports replica behavior, not direct observation of medieval
technique.
[KN3] Hoshino et al. (1990), part 2, 竹中大工道具館研究紀要 Vol. 2. DOI 10.50862/dougukan.2.0_1.
— kind: museum-bulletin research/technical report. Second-stage tooth-angle
experiment; describes the Kusado saw as regarded as Kamakura-period; its conclusion
says it established only basic parameters and expects further study [review, relayed].
Not read this pass.
[KN4] Hoshino et al. (1991), part 3, Vol. 3. DOI 10.50862/dougukan.3.0_57.
— kind: museum-bulletin research/technical report. 55° isosceles 箱屋目 test tooth;
four nageshi angles; proposed 35–40° optimum for that configuration; no significant
difference among the four saws [review, relayed]. Not read this pass.
[KN5] Watanabe Akira (1999), "古代・中世における建築用主要道具について," Vol. 11.
DOI 10.50862/dougukan.11.0_1. — kind: museum-bulletin research/technical report.
Inferred (推定される) medieval generalization of pull use; concave-to-convex tooth
line in antiquity as inference [review, relayed]. Does not say the konoha-gata family
"converged" on pull or that maebiki push is a survival. Not read this pass.
[KN6] Razorsaw / Gyokucho (玉鳥産業), 鋸の歴史・年表. https://www.razorsaw.co.jp/support/rekisi.html
— kind: manufacturer history (MEDIUM).
[KN7] Wikipedia (JA), 鋸. https://ja.wikipedia.org/wiki/鋸 — kind: encyclopedia. Load-bearing
historical claims drawn from it (Kofun micro-saws, Ishiyama-dera depiction) are
graded no higher than MEDIUM and should be traced to the published reference it
cites (『図説世界の木工具事典』, 2015).
[KN8] Wikipedia (JA), 番匠. https://ja.wikipedia.org/wiki/番匠 — kind: encyclopedia (MEDIUM);
background only.
[KN9] Wikipedia (JA), 鋸 (重要文化財工第922号): Hōryū-ji treasure saw, TNM N-141 — kind:
encyclopedia (MEDIUM).
[KN10] Kasuga Gongen Genki E (1309): en.wikipedia scroll identity; kabumoku.exblog.jp firsthand
exhibition account — kind: encyclopedia + blog (texture; LOWER for any saw claim).
[KN11] Working by Hand, "Exploring Japanese Saws."
https://workingbyhand.wordpress.com/2024/02/23/exploring-japanese-saws-iii-saw-types/
— kind: blog (LOW).
[KN12] JAANUS, nokogiri 鋸. https://projects.mcah.columbia.edu/jaanus/node/2011 — kind:
academic reference (general class norms only).
[KN13] Yoshikawa Kinji, 『鋸』, 法政大学出版局 (1967/1976) — kind: book; cited by [KN2]; not read.
[KN14] USDA Forest Service / PCTA, Crosscut Saw Limbing & Bucking (Module 3.2) and Crosscut Saw
Manual — kind: technical guide. Applies to loaded logs and limbs; NOT evidence for
konoha-gata-specific hazard. Not re-read this pass.
[KN15] Hiroshima Prefectural History Museum, Kusado Sengen site collection.
https://www.pref.hiroshima.lg.jp/site/rekishih/materials.html — kind: museum. Supports
the site date range and collection; not the saw's accession number, layer or date
[review, relayed].
[KN16] Covington & Sons, "Japanese Handsaws: The Maebiki Ooga." — kind: blog (general laminate
norm; conditional only).
[KN17] Takenaka exhibition via Japan House, anahiki/ibarame descendant anatomy
https://www.japanhouselondon.uk/read-and-watch/the-tools-used-in-the-craft-of-carpentry/
— kind: museum exhibition (descendant contrast only).
[KN18] Akanuma & Fukui (1997), "Saw use in Japan before the modern period," 竹中大工道具館研究紀要
Vol. 9. https://www.jstage.jst.go.jp/article/dougukan/9/0/9_BTCTM0902/_article/-char/ja/
— kind: museum-bulletin research/technical report. Tooth-form change toward pull-use
around the fifteenth century; clear material distinction then; pictorial confirmation
of pull-use from the seventeenth century; does not establish descent from the Kusado
specimen [review, relayed]. Not read this pass. (Replaces the undefined
[KN-via-Anahiki].)
[KN19] Hoshino et al. (1990) part 2, same as [KN3], cited for the "regarded as Kamakura-period"
attribution [review, relayed].
[KN20] Hiroshima Prefecture, designated Kusado Sengen objects catalogue (wood objects).
https://www.pref.hiroshima.lg.jp/uploaded/attachment/591444.pdf — consulted by the reviewer as
a lead; not the iron saw [review, relayed].
```

The reviewer's corrections (11)

• SHOULD-FIX — 1. Stop presenting the reconstruction series as peer-reviewed — Exact claim: “peer-reviewed experimental reconstruction,” “three-part peer-reviewed study,” and “peer-reviewed museum bulletin.”
• BLOCKER — 2. Recast the mixed-stroke result as replica-test behavior, not the tool family’s origin story — Exact claim: “the deep crosscut ancestor was NOT a pull saw,” “the family may have begun mixed-stroke and converged toward pull,” “The konoha-gata is neither—it is the mixed-stroke ancestor,” and the beginner warning that pull-only work “risks losing half its stroke and most of its efficiency.”
• BLOCKER — 3. Remove the unsupported maebiki-survival claim — Exact claim: “The maebiki’s push is a survival of the ancestor’s stroke, not a foreign intrusion,” and “may be better understood as a possible survival of the ancestor’s mixed stroke.”
• SHOULD-FIX — 4. Narrow “ancestor,” “common ancestor,” and direct-descendant language — Exact claim: “deep ancestor of the whole Japanese crosscut line,” “the common ancestor,” “shape-parent of the Anahiki,” and “the single, general-purpose leaf saw ... was gradually displaced by the purpose-built saws.”
• SHOULD-FIX — 5. Downgrade “securely dated” and identify the artifact’s actual dating basis — Exact claim: “an excavated, dated medieval specimen,” “securely-dated excavated object,” and the broad parenthetical “occupied ~Kamakura–Muromachi, 13th–16th c.” used as the object’s date.
• SHOULD-FIX — 6. Correct the plate-thickness threshold and keep every dimension experiment-scoped — Exact claim: “~1.1 mm minimum plate thickness,” “a plate of 1.1 mm or less is not practical,” and Image 3’s “~1.1 mm minimum thickness.”
• SHOULD-FIX — 7. Soften absolute job exclusions and avoid turning a test result into universal practice — Exact claim: “Ripping ... never its job,” “never the leaf saw’s,” “The konoha-gata is neither,” and “worked best ... in short localized strokes.”
• BLOCKER — 8. Rebuild the safety section around evidence appropriate to this tool and scale — Exact claim: “The governing hazard is the general crosscut/bucking class: a kerf closing under compression ... tension wood separating violently,” followed by western bucking guidance; plus “a kink ... is not straightenable to tolerance without specialist hammer-work.”
• SHOULD-FIX — 9. Remove unsupported terminology and named-hands deductions — Exact claim: “The 「いわゆる」 (‘so-called’) signals that this is a scholarly descriptive name, not a period term—the medieval users had no single recorded word for it,” and “none is expected ... a distinct saw-smith trade had not yet crystallized.”
• SHOULD-FIX — 10. Repair internal contradictions, broken references, and workflow residue — Exact claim: The main prose labels the stroke genealogy uncertain, while ## Images Needed instructs Forge to caption it as settled; the text uses [KN-via-Anahiki]; the file ends with </content>; provenance narrates “the model,” “the model,” research passes, and a prior apply pass; frontmatter status embeds a …
• SHOULD-FIX — 11. Split confidence by claim instead of granting HIGH to the whole package — Exact claim: “HIGH for WHAT THE TOOL WAS, WHEN, and HOW IT WAS USED,” and “the two genuinely-new structural findings ... are HIGH.”