c.1325, "snowflake, spark of fire," from O.N. flaga "stone slab, flake" (see flagstone); sense of "defect, fault" first recorded 1586, first of character, later (1604) of material things; probably via notion of a "fragment" broken off.
Working in technical support, the subject of flaws often is discussed... flaws in software ("bugs"), flaws in hardware, flaws in documentation, etc.
I was thinking about this word the other day and realized that, in English, we probably use it equally in either a literal sense or figurative sense -- a character flaw (of a person), a facial flaw (of a sculpture), or a flaw in a plan (which is literal, in a sense, but more metaphorical since it does not relate to a physical thing).
When I checked the trusty Online Etymology Dictionary, I was intrigued by the relationship between flaw and scar:
1388, from O.Fr. escare "scab," from L.L. eschara, from Gk. eskhara "scab formed after a burn," lit. "hearth, fireplace," of unknown origin. Eng. sense probably infl. by M.E. skar (1390) "crack, cut, incision," from O.N. skarð, related to score. Fig. sense attested from 1583. The verb is first recorded 1555.
O.E. sceran, scieran (class IV strong verb; past tense scear, pp. scoren), from P.Gmc. *sker- "to cut" (cf. O.N., O.Fris. skera, Du. scheren, Ger. scheren "to shear"), from PIE *(s)ker- "to cut, to scrape, to hack" (cf. Skt. krnati "hurts, wounds, kills," krntati "cuts;" Hittite karsh- "to cut off;" Gk. keirein "to cut, shear;" Lith. skiriu "to separate;" O.Ir. scaraim "I separate;" Welsh ysgar "to separate," ysgyr "fragment").
O.E. sceard "fragment, gap," from P.Gmc. *skardas, a pp. from the root of O.E. sceran "to cut" (see shear). Cf. Du. schaard "a flaw, a fragment," Ger. Scharte "a notch," Dan. skaar "chink, potsherd." Meaning "fragment of broken earthenware" developed in late O.E. Fr. écharde "prickle, splinter" is a Gmc. loan-word.
And I was even more surprised at the relationship between flaw and fragment that was revealed in all those related words!
...not to mention the possible relationship to harsh!
originally of texture, "hairy," 1533, probably from harske "rough, coarse, sour" (c.1300), a northern word of Scand. origin (cf. Dan. and Norw. harsk "rancid, rank"), related to M.L.G. harsch "rough, raw," probably from PIE base *qars- "to scrape, scratch, rub" (cf. Lith. karsiu "to comb," O.C.S. krasta, Rus. korosta "to itch," L. carduus "thistle," Skt. kasati "rubs, scratches").
The PIE base here makes me wonder if there may be more to the English usage of "curse" meaning "harsh words"!
Note: I'm sure there are PIE literalists who will knock this hypothetical thinking... but I also don't believe anyone who says that PIE root *qars- cannot be related to PIE root *(s)ker-.