See this lexicum : https://awb.saw-leipzig.de/?sigle=AWB&lemid=D01106 It seems clear in the case of Old High German that farmstead and farmland are the first meaning.  And it is even clearer that Old High German retained different meaning and different expression related to this root suggesting a general shift from the singular to the collection. The direction of the change is obviously from a singular farmstead to a collection.  The entry breaks down the definition into four distinct historical layers: 1. The Estate or Farmstead (Hof, Landgut) In this earliest sense, thorf refers to private property—essentially a "manor" or a "ranch."     a) The Reluctant Guest: One citation describes a man excusing himself from a banquet: "thorph coufta ih... inti gisehen iz" ("I bought a farm/estate... and I must go see it"). This uses the Latin villam as a reference, meaning a private country estate.     b) The Prodigal Son: In the famous biblical parable, the master "santa inan in sin thorf, thaz her fuotriti suuin" ("sent him to his farm/estate to feed pigs"). Here, the thorf is clearly a specific agricultural property owned by an individual. 2. The Village or Rural Settlement (Siedlung) This is where the word starts to describe a collection of houses, often defined by what it isn't (i.e., it isn't a fortified city).     a) The Unwalled Town: The text defines dorf as a "vicus"—a place that has streets but "sine muris" ("without walls").     b) The Contrast with Cities: One example says, "manige uuesen in demo dorf, unmanige in dero burg" ("many people are in the village, but few are in the [fortified] city"). This shows dorf becoming a category of settlement size.     c) Specific Locations: The text mentions "thaz thorf thaz dar giquetan ist Gethsemani" ("the village that is called Gethsemane"). 3. The Neighborhood or Quarter (Stadtviertel) This is a more niche use where thorf describes sections of a larger urban area, often in the plural (thorphun).     a) Public Display: A warning against hypocrisy: "so thie lihhazara tuont in dingun inti in thorphun" ("as the hypocrites do in the assemblies and in the [streets/neighborhoods]"). Here, it translates the Late Latin vicos, meaning the public blocks or quarters of a town where people gather. 4. The "Crowd" or "Peasantry" (Menge von Bauern) This is the most "abstract" evolution. The word for the place starts being used to describe the type of people found there.     a) The Rustic Multitude: One entry notes a scribe used dorf to translate rusticam manum ("a rustic hand" or "a crowd of country-folk").     b) The "Village-Dweller": The text even mentions a derivative: uillanus dorphere (literally a "villager" or "peasant"), showing how the word was used to build new social categories. Old High German retained the direction of this change in various ways. There is no doubt about that. The Gothic Bible uses the word haims to translate the Greek kōmē (village). Place names ending in -heim (in Germany) and -ham (in England) appeared frequently during the Migration Period (5th-6th centuries).  There are also runic inscriptions that make more sense if the root is interpreted as meaning village or community. At least the shift looks more ancient.  The codex Abrogans is a glossary, for example it is simply written the equivalence between 'uilla' (villa) and 'thorf'.

@Genava55 The Codex Abrogans is from the 8th century. I shortened “villa worker” to “villager”, maybe changing the exact meaning, but the point remains that these things happened, even when they are later examples. I’m just trying to make sense of what the dictionaries actually say. Another issue could be how big these settlements are in context. You say the original meaning is farmstead, but what if a very small collection of farmsteads is also valid? At Proto-Germanic times this would have been the equivalent of a village (again, there has to be a reason why this has been reconstructed as such, besides commonality), and what meaning exactly had on the different branching languages could have turned out to be quite arbitrary and relative, Gothic could have retained a more farmstead position, while all the others considered the increased demographics. And related to this, coming back to the very beginning, what does say that þurpą has to be smaller than haimaz, which is primarily reconstructed as home? Maybe you can quote the relevant passages from the Codex Argenteus and Abrogans.

Again, as I said, in Codex Abrogans the meaning leans more towards a farmstead for Old High German. It is only later that the semantic shift happens for this branch. Your example with 'villain' is not correct. 'Villanus' doesn't mean 'villager' nor is equivalent for 'villager'. A villanus is a peasant or a worker in a farm estate, a villa rustica in the Roman perception. 'Village' is a semantic shift appearing much later during the medieval period and once again in the same direction: from a single farmstead to a collection of farmsteads. Villanus means 'villager' only in Middle English after a borrowing from French, where the semantic shift already happened. I don't see anything convincing for the moment to change my opinion.

@Genava55 But Euler’s work comes after Ringe’s. Downgrading happens particularly with things perceived as "less than": ‘villain‘ comes from Latin ‘villanus‘, meaning ‘villager‘ (which would give credence to ‘village‘ being downgraded to ‘farmstead‘ and not the other way around), or ‘sinister‘, coming from Latin ‘left-handed‘, obvious for Italian speakers. So, what happened? A downgrading for Gothic or a collection for West and North Germanic languages? Even if Gothic was the more conservative, a consensus considering many other variables was reached and written down in Proto-Germanic dictionaries. If we grab a time machine we might find out that this was the wrong call, but it wouldn’t be an error on our part, but the academic consensus, and what can one really do about that but guess.

Euler and Maurer are certainly respected, but they represent the 'divergence' school of the mid-20th century. Modern computational phylogenetics (like the Ringe work I cited) suggests a much tighter 'Germanic Core' than Maurer’s 1940s theories allowed for. Regarding the 'Elephant in the Room': Gothic is likely conservative because of the direction of semantic change. It is a documented linguistic trend for words describing land or enclosures to expand into words for settlements (think of English 'Town' from *tuną 'fence/enclosure'). It is almost unheard of for a word meaning 'village' to suddenly be downgraded to mean 'dirt field' in a single branch. As for why dictionaries list it as 'village': Dictionaries are tools of commonality. Since the West and North Germanic branches (which cover 90% of our surviving texts) use 'village,' dictionaries list that for the sake of the user. But if you look at the etymological root *treb-, the meaning is clearly 'to build' or 'a dwelling.' A village is a collection of dwellings; a farmstead is a single dwelling. Gothic, being the earliest snapshot we have, simply catches the word before the 'collection' phase took over the West Germanic dialects.