understand
To understand is to grasp a meaning — and the folk reading of the word is that you get beneath a thing, stand under it, support it from below. The scholarship says probably not. The leading account (Barnhart's) reads the under of understand not as "beneath" but as an old, homophonous under meaning "between, among" — the same root as Latin inter, Sanskrit antar, Greek entera ("the inner parts"). On that reading, to understand is to stand in the midst of a thing, to be positioned among its parts rather than below the whole. The reading is contested, and genuinely unresolved — Old English ran the two senses of under together, and modern under is itself a merger of two roots — but the leading account holds that comprehension was first imagined not as depth but as standing among.
The natural reading of understand is that you stand under a thing — you get beneath it, support it, hold it up from below, and that is what it is to grasp it. It is a satisfying picture and it is probably wrong. The leading scholarly account takes the under here to be a different word from the under of undergo and undermine. Old English had, alongside the spatial under ("beneath"), a homophonous under meaning "between, among, in the presence of" — and that second under is on the root *h₁entér, the same root as Latin inter, "between." On this reading, to understand is to stand among: to take up a position in the midst of a thing's parts, rather than below the whole of it. Comprehension as location, not as depth.
You can see the second under elsewhere in Old English, in a little family of verbs where the prefix plainly means "perceive" rather than "beneath": underfindan, to be aware; undersecan, to examine. And the typological company is good — Greek did the same trick with a stand-verb, building epistamai, "I understand," out of "I stand upon." Across languages, the act of comprehension keeps getting imagined as a body taking up a stance with respect to a thing. The English word picked the stance "among"; the Greek picked "upon"; the folk etymology picked "beneath." Only one of the three is likely to be the real history, and it is not the intuitive one.
This is an honest place to flag the uncertainty, because the word resists a clean story. Etymonline presents the "among" reading as Barnhart's suggestion and then notes that other authorities keep "among, between, in the presence of" as simply another sense of the ordinary word under — not a separate root at all. Worse for tidiness: Wiktionary derives the modern English preposition under from a merger of the two PIE roots, *ndʰer- ("beneath") and *h₁entér ("between") — so the two senses are not cleanly separable in the one English word that carries them both. "The under in understand is a totally different word" overstates it. The careful claim is narrower and stranger: there were two unders, they have been bleeding into each other for a thousand years, and the leading account holds that the one inside understand began as "among" — though even that is a reconstruction, not a settled fact.
The false friend is the rest of the under-words. Undergo, undertake, undermine, understudy — these carry the literal "beneath" under, from *ndʰer-, the root that gives Sanskrit adhah, "below," and is cousin to Latin infra. They look identical to the under of understand and, on the leading account, are not the same word. The -stand is real and shared — it is the ordinary stand, PIE *steh₂-, the root of stable and statue and stage. It is only the first syllable that hides the question. To understand something may be less like getting under it than like standing inside it — surrounded by its parts, in the middle of the thing, close enough to be among.
- inter- (international, intervene, interval) — Latin inter, "between, among," from the same PIE *h₁entér as the proposed under of understand. Etymonline lists Old English under and Gothic undar as cognates of inter under this root — the strongest single confirmation of the link
- antar / antara — Sanskrit, "within, between, interior"; the same *h₁entér root, in the Indo-Aryan branch
- entera / enteric / dysentery — Greek entera, "the inner parts, the intestines" — literally "the things between/inside"; the same root, naming the body's interior
- stand / stable / statue / stage / stem — the second element, -stand; all from PIE *steh₂-, "to stand, be firm." Cousins of the back half of the compound, not the front
- epistemology — Greek epistēmē, "knowledge," from epistamai, "I understand," literally "I stand upon" — a parallel stand-verb for comprehension, by a different preverb
- under (the ordinary preposition) — part-cousin — modern English under is itself a merger of two roots: the "between" *h₁entér (cognate with understand's under) and the "beneath" *ndʰer-. Related on one strand only, which is exactly why the etymology is contested
- undergo / undertake / undermine — different root (on the leading account) — these carry the "beneath" under, from PIE *ndʰer- (Sanskrit adhah, "below"; cf. Latin infra). Same spelling as understand's under, but — if Barnhart is right — a different word
- inter (to bury) — different root — from Latin in- + terra, "to put into the earth" (PIE *ters-, "dry, earth"). It looks like the prefix inter-, which is a true cognate, but the burial verb is unrelated — a double trap
- "stand under" (the folk reading) — folk etymology — the satisfying gloss "to understand = to stand under / support" has no separate root; it is a reanalysis of understandan with under read as "beneath." It feels explanatory and is the reading the scholarship argues against