Any designer who has ever pressed a swatch between thumb and forefinger knows that fabric hand is a language without a dictionary. We say “soft,” “crisp,” “mealy,” “springy,” but those words drift across cultures, mills, and hands. At Vivina, I treat fabric as seriously as silhouette — and that demands a rigorous, shared vocabulary for touch. The Kawabata Evaluation System (KES) provides exactly that: an objective measurement framework that translates mechanical and surface properties into standardized sensory descriptors. Here, I want to walk through the primary hand equations that form the backbone of the system, unpack their sensory logic, and build a taxonomy that anchors each descriptor — koshi, numeri, fukurami, shari — in the material choices that generate it.
The KES-F instrument suite measures a fabric’s response to low-stress deformations, mimicking the forces a hand actually applies when assessing drape or softness. It captures sixteen parameters, later transformed into primary hand values via multiple linear regression. These parameters fall into six blocks: tensile (LT, WT, RT, EMT), bending (B, 2HB), shear (G, 2HG, 2HG5), compression (LC, WC, RC, T0), surface friction (MIU, MMD), and surface roughness (SMD). Each number is a physical quantity — bending rigidity B in gf·cm²/cm, coefficient of friction MIU, mean deviation of surface roughness SMD in microns — giving a precise tactile fingerprint that no hand alone can quantify.
Expert panels then rate the same fabrics on a set of Japanese hand descriptors on a 0–10 scale. The KES methodology uses stepwise linear regression to derive equations that predict a hand value (HV) for a given descriptor from a subset of the mechanical parameters. The general form is:
HV = C₀ + Σ Cᵢ · Xᵢ
where Xᵢ are either the raw parameters or logarithmic transformations (selected for distribution shapes and physical linearity), and Cⱼ are coefficients fitted to the expert consensus. While the exact coefficients are tuned to Japanese fashion norms and are part of the proprietary KES standard, the underlying causal structure is well-documented: each primary hand equation selects a few dominant parameters, and then adjusts for interactions.
Take koshi — stiffness, the feeling of a fabric that resists bending and shearing with a "brittle" snap. Its regression typically leans on bending rigidity B (high), bending hysteresis 2HB (low for recovery), shear stiffness G (high), and sometimes weight per unit area W. A crisp cotton lawn or a starched linen shows high koshi. Numeri, on the other hand, describes a smooth, slippery, almost moist sensation — think silk charmeuse or a glazed chintz. Its equation hinges on low surface friction MIU, low friction variation MMD, and low roughness SMD. These parameters are direct consequences of fiber shape (round, smooth silk vs. scaly wool) and finishing (calendering, mercerization).
Fukurami — a bulky, springy fullness you feel when you squeeze a fabric and it pushes back with soft volume — is driven by compressional properties: a high linearity of compression LC (the fabric resists early and elastically), a high compressional energy WC (it takes work to compress), and a substantial thickness T0. Brushed wool flannel or a lofty down-proof nylon are textbook examples. Shari, the crisp rustling hand of a cool, dry fabric like ramie or high-twist cotton voile, comes from high bending rigidity B combined paradoxically with low surface friction MIU — it feels stiff yet sleek, the fibers sliding against each other rather than catching. Fabric construction here is key: long-staple fiber, high twist, and a dense plain weave.
These four — koshi, numeri, fukurami, shari — are the heart of the KES primary hand for many apparel end-uses, but the system includes others. Hari (anti-drape stiffness, like a spreading skirt) is related to koshi but weighted toward bending and weight. Kishimi is the scrooping, slightly creaky feel of certain silks; its equation detects a specific pattern of surface friction and bending recovery. Shinayakasa (flexibility with a soft, supple touch) is essentially the inverse of stiffness, drawing on low B, low G, and high compliance. Each descriptor has its own regression model, validated on hundreds of fabrics, with R² values often exceeding 0.8, meaning the instrument panel replicates the expert hand panel with remarkable fidelity.
Once you have the primary hand scores for a fabric, KES then calculates a total hand value (THV) — a single number from 0 to 5 that combines the relevant primaries for a given end-use (e.g., men’s suiting, women’s dress, or sportswear). But for me, the real power is not in the final score but in the map it draws between physical cause and sensory effect. Knowing that a fukurami value can be pushed up by switching from combed cotton to a hollow-core polyester, or by raising the fabric with a teasel, means you can design hand, not just inherit it from a mill.
This is the taxonomy I’m building into my work at Vivina: a structured bridge from fiber → yarn → weave → finish → mechanical parameter → primary hand → the words a designer will actually use. When I examine a deadstock cotton plain-weave under the KES lens, I see low B, moderate G, a medium MIU, and a crisp SMD — and I can say “this will feel soft and slightly crisp, with a fine granular surface; it’s a canvas for enzyme washing to enhance numeri.” That’s not guesswork; it’s the sensory logic of the primary hand equations made routine. As more researcher minds join our studio — focused on knits, sustainability, regional markets — I plan to extend this taxonomy to cover handle variations across stretch fabrics, pile constructions, and coated surfaces, always grounded in the same mechanical first principles. Because in the end, a fabric’s hand is not a mystery; it is a set of physical truths waiting to be read by the right instrument, and by the hands that are learning to speak its language.
Comments