Anatoly at Avva has an interesting post; I’ll translate from his Russian:
1. Why exactly is it “часовые любви” (‘the sentries of love’ [title of a Bulat Okudzhava song]) but “квартира Любови Павловны” (‘the apartment of Lyubov Pavlovna’)? What caused this difference in declension [genitive lyubvi vs. Lyubovi], how did it develop?
2. Is there a name for this phenomenon (what is initially the same word gets declined/conjugated differently depending on the meaning), and what other interesting examples are there?
One of his commenters says that the word Любовь as a female name must somehow be distinguished from любовь as a feeling, but this of course is simply a rationalization parallel to “we have to spell its and it’s differently to avoid ambiguity.” At any rate, I thought the question about other examples (in languages that decline their words) was worth thinking about.
Burned/burnt and similar dual past participles (I think there are a few though I can’t remember any more offhand)
Without buying into “must somehow be distinguished,” maybe the key phrasing is “initially the same word,” with the emphasis on “initially.” Are Любовь and любовь currently just different senses of “the same word”? I dunno, but not obviously.
An English example might be “dwarves” versus “dwarfs” as the plural of “dwarf,” with the attempt to harmonize them by saying each is correct (and allegedly the only correct form …) for particular senses of “dwarf.” Tolkien may also be responsible for elven versus elfin.
An English example might be “dwarves” versus “dwarfs” as the plural of “dwarf,” with the attempt to harmonize them by saying each is correct (and allegedly the only correct form …) for particular senses of “dwarf.”
That’s a good example. But burned/burnt and the like are just alternate forms with no distinction in meaning.
English “hang” → “hung”/“hanged” was the first that came to mind for me.
It took me a while to remember the famous baseball example, where the past tense of the phrasal verb “to fly out”* is “flied out” rather than “flew out,” with the “flied” form attested as early as 1872.
*A gloss for the non-baseball-conversant foreigners that I realize may still be opaque: “To hit a fly ball that is caught for an out before it hits the ground.”
But burned/burnt and the like are just alternate forms with no distinction in meaning.
I’d say there is. Burnt toast is on the dark side. Burned toast is inedible charcoal.
Re burned/burnt: some have observed that for some speakers (esp. Americans) “burnt” can be used as the adjectival form of the participle but not as the participle proper actually used in a verbal construction. Similarly, some might use adjectival “spilt” in the fixed phrase “spilt milk” but wouldn’t say “the toddler had spilt the milk onto the kitchen floor.”
Arabic has a few cases like that, where the proper name is diptotic but the common noun from which it derives is triptotic: miṣr-un/-an/-in ‘border, large settlement’ vs. Miṣr-u/-a/-a ‘Egypt’ (nominative/accusative/genitive respectively.)
I feel sure some (former) Surrey Morphology Group stalwarts must have written about this for Russian, but I can’t think of a paper offhand.
Thanks for the Arabic — that’s exactly the kind of thing I had in mind.
Not declension, but perhaps somewhat analogous given the way that Volta-Congo noun class affixes are the method for number-marking on nouns: the way that Volta-Congo languages switch a stem from one noun class to another to produce a meaning difference, like
Swahili m-toto “child”, ki-toto “small child”,
(wa-toto “children”, vi-toto “small children”),
Mbelime cīɛ̀-n-ú “road”, cīɛ̀-kɛ̀ “path”,
(cīɛ̀-n-í “roads”, cīɛ̀-sí “paths”),
Mbelime bíí-kɛ̀ “child”, bíí-hũ̀ “fat child”,
Kusaal pusi-g “tamarind tree”, pusi-r “tamarind fruit.”
In German, senden has two sets of past forms, depending on whether what is being sent is or is not in the form of a radio (or television) signal.
A bit specialized: English data / datums.
Less so: media / mediums.
Also: fish / fishes.
A South African participant in alt.usage.english named Steve Hayes said that for him, “learned” was temporary, maybe not beyond the final exam, but “learnt” was permanent.
I suspect you can say that differences in meaning between such forms are idiosyncratic.
A pretty consistent one in English is “indexes” of books, “indices” of matrix components etc. After a look at the OED, I’m not sure there’s any reasonably consistent one for “matrixes” and “matrices”.
A personal-name example in Kusaal is Akudug “Akudugu”, which literally means “(Iron) Nail”; as a common noun, kudug has been completely ousted by its original plural form kut, which has been duly kitted out with a new plural form, kutnam. One nails, two nailses.
Thanks for the Arabic — that’s exactly the kind of thing I had in mind.
Is this close to what you had in mind? “There were two Pattys cooking patties in that restaurant.”
And now for examples that are different from what you had in mind. Spanish has two deverbal nouns from esperar: espera for a wait and esperanza for (a) hope.
For the opposite pattern, “bases”, with different pronunciations, is the plural of both “base” and “basis”. In Spanish, the past tenses of ser and ir are identical (fui, fuiste, etc.), but the other parts of the verbs are quite different. Some weird suppletion?
ETA: As the French detective said, “Je suis, donc je suis.”
Born–borne
Struck–stricken
Littlest–least
Baddest–worst
Slayed–slew
1syll – 2 syll
Learned
Crooked
Cleft–cloven
Fish–fishes
Beefs–beeves
Distinguished senses for some speakers:—
Shined–shone
Pleaded–pled
Speeded–sped
Former inflections now separate lexemes include:—
past, covert, wrought
There should be lists somewhere
—-
ETA
Ooh the foreign-v-native plural distinction is fruitful: cherub, stigma,
That may be so for some speakers, but it mostly depends on where you’re from.
“Related” is always verwandt (and “relatives” are Verwandte), but “used” is verwandt and/or verwendet depending on geography. (Exclusively verwendet for me, as you’d expect… fuck. Yet another paper, or book chapter in this case, that has disappeared from academia.edu. It’s still in the list on the author’s page, and there’s still a download link for those foolhardy enough to create an account and receive spam forever, but reading it on the site seems no longer possible. The second volume of the book it’s in is in Google Books, but the chapter is in the the first volume, which lacks a preview in all its numerous copies, and the chapter isn’t even uploaded on the author’s ResearchGate page.)
I have an account with a.e and I don’t receive spam from them. All I had to do was go to the settings page and check off about a godzillion boxes corresponding to different categories of such.
I don’t know how widespread it is, or used to be, but I’ve read Norwegian dialect grammars that describe neuter words with different plurals for different meanings. Typically, the basic or original meaning will keep the regular neuter plurals, while the new or extended meaning will take feminine plurals. The example I remember now is:
bord n. “plank” – ¹borde – bord – ¹borda
bord n. “table” – ¹borde – ²border – ²bordene
(In your average Eastern dialect, <bord> can be phonemized /bu:ɽ/.)
I suspect subtle borrowing in a specific sense — or lexicalized codeswitching, if you prefer. Traded objects were more likely to be spoken of in a “Riksmål” (H) register with regularized paradigms.
Edit: The variation in plural forms is obvious to anyone listening to colloquial Norwegian. What the dialect grammars made me realize was that this could be systematic rather than idiosyncratic.
The plural of Latin locus is loca (irregular!) and loci, depending on the meaning. But if you look into a good dictionary you’ll find that even among classical writers (i.e. Caesar and Cicero) the usage difference is not as clear-cut as school grammars pretend.
senden/sandte in the sense of “sending a letter/a package/amessenger” has been largely replaced in everyday language by schicken. senden/sendete in the sense “broadcast” is still part of everyday language. The noun Versand is still a part of everyday language, though.
Norwegian dialect grammars that describe neuter words with different plurals for different meanings
Welsh, too, has a few nouns where the plural distinguishes two meanings which fall together in the singular, like llwyth “tribe”, plural llwythau, llwyth “load”, plural llwythi; pryd “meal”, plural prydau, pryd “time”, plural prydiau; so too personau “persons”, personiaid “parsons”, though the singulars also differ in grammatical gender in that one. [Same with brawd “brother”, plural brodyr, versus brawd “judgment” (feminine), plural brodiau, where the resemblance between the singulars is just accidental.]
I suppose you could make a case that in modern English “brother”, plural “brothers” versus “brother” pl “brethren” is something similar.
What about “ring, rang, rung” versus the unrelated “ring, ringed, ringed”?
J.W.’s “flied out” reminded me of weightlifters’ “flys” or “flyes”, exercises where one moves the straight arms as if flying, not to be confused with the various meanings of “flies”. Edit: The singular is often “flye”, so this is getting too complicated.
English “brother”, plural “brothers” versus “brother” pl “brethren” is something similar.
There’s also ‘bro’ plural ‘bros’. I’m not hip enough to be sure, but the singular can include ‘male sibling’; whereas ‘bros’ I think means ‘mates’ (or the derogatory meanings), can’t mean biological brothers.
Hebrew: עֲדָשָׁה ʿădāšâ (f.), ‘lens’ pl. עֲדָשׁוֹת ʿădāšôṯ, ‘lentil’ pl. עֲדָשִׁים ʿădāšîm.
Others with different plurals are of different genders so arguably not the same (עֶצֶם ʿeṣem, f. ‘bone’, m. ‘thing, material object’), or are parts of fixed expressions. I can’t think of any such pairs where both are biblical.
The difference there is purely one of spelling though and this difference may have arisen simply to distinguish the two senses cf. German dass and das.
“Born–borne The difference there is purely one of spelling though”
Nope; some of us still have the FORCE–NORTH split. Still relevant in any case.
And speaking of homonyms, -graphs and phones…
What about “ring, rang, rung” versus the unrelated “ring, ringed, ringed”?
if you are moving beyond polysemy, you might add “wring–wrung–rung”.
the past tense of the phrasal verb “to fly out”* is “flied out” rather than “flew out,”
That depends on what the meaning of “is” is. “Flied out” is universally prescribed, yet “flew out” persists in speech and writing, including professional sportswriting: e.g. about a dozen within the last six months in the New York Times, most recently on Oct. 16 (“They groaned when Cal Raleigh flew out to foul territory”). See Language Log (2012) and Language Log (2007) for some statistics — the “flew/flied” ratio in “fl? out to ___ field” can be as high as 1/2 or more in some corpora.
In RP which uses the split born can be /bɔːn/ or /bɔːrn/ and borne can be /bɔːn/ or /bɔːrn/.
So, even with the split, the two remain homophones. In certain conservative Scot dialects or in hypercorrect RP the vowel in borne might be slightly lenghtened but that’s about it. Of course, born and borne are more than just virtual homophones, they are in etymological terms exactly the same word.
Even the semantic distinction is rather one of convention, not one of substance and that’s why the two words are not always clearly distinguishable semantically. In Mortal Kombat Deception which I used to play with my grandson back in 06 the intro ends with the sentence: ‘Only I can destroy this threat, borne of deception’. When I first heard it I remember wondering if I should interpret this as ‘generated by’ (along the lines of birth) or ‘carried by’ (as in to bear) but the meaning is the same in both cases.
My grandmother went through many births. She didn’t remember exactly when her kids were ‘born’, but could vaguely recall a few things about the period they were ‘borne’ in her belly, especially the period of the third semester when mobility is most noticeably affected and the feeling of ‘bearing’ something is at its heaviest. I might be wrong but I have the inkling that in older times rural people didn’t closely track their birthdays but only had a vague sense of the time they were ‘born’ based on events that happened when they were being carried. So originally ‘I was born’ meant ‘I was carried’ not ‘I came to the world’ (zur Welt kommen) as we usually put it in German.
even with the split, the two remain homophones
RP does not have the split. I do have the split and I can assure you that born is NORTH and borne is FORCE. If you don’t believe me, here‘s John C. Wells Accents of English Volume 1 page xix
Other English differences are auxiliary- vs main-verb non-assertive “do”, “have”, and “need”, and the (non-New Yorker) AmE distinction between “have got” and “have gotten”
I stumbled across an example in this comment I left four years ago. If you are talking about crustaceans, the normal plural is the zero plural shrimp. However, if you are talking about two short guys, it has to be shrimps.
I did that a while ago, and I get a lot less spam than before; but that wasn’t enough.
Agreed on schicken and Versand; would not have guessed that born is NORTH or that borne is FORCE (because the RP-oid sound system I was sort of taught has the merger) and can’t figure out how that happened, but it’s probably not any stranger than forty.
@At any rate, I thought the question about other examples (in languages that decline their words) was worth thinking about.
Here are some examples from Polish, where there is a distinction between plural collective noun referring to the country name and plural form referring to people of that country:
Niemcy – Germany (country) vs. Niemcy – nominative plural of Niemiec – Germans. These words have different forms for genitive, accusative and locative
Czechy Czechia (country) vs Czesi – Czechs. The dative, instrumental and locative forms are the same.
Węgry – Hungary vs Węgrzy – Hungarians. Dative and instrumental the same.
Włochy – Italy vs Włosi – Italians. Dative and instrumental the same.
Thank you for those much needed corrective comments, mollymooly, reminding us of the diversity and history of English!
On the history of the vowels in born and borne, see for example especially §13.353 in Jespersen, A Modern English Grammar On Historical Principles, part I, here, along with the previous sections §13.351–352. On the way the spelling of these forms shook out, see also the note in the OED under bear (available for all here on archive.org, section 44, p. 733). Short comment because I have to go to bed now.
@DM: I have the impression that most words spelled with a silent vowel after the “o”, such as “force”, are FORCE, and those without such a vowel, such as “north”, are NORTH. Awaiting corrections from people who actually know.
If like me you spent much of your teens being exposed to American rock music recordings released in 1975 (the year I had turned 10 and wasn’t yet listening to either, but they lingered …), there was the minimal pair of “Born to Run” versus “Borne on Wings of Steel.”* Which differed in various other ways, including stylistic subgenres, but still.
*Full title is longer, but I wanted to make the contrast clean and you would know what I meant if you knew what I meant.
Niemcy – Germany (country) vs. Niemcy – nominative plural of Niemiec – Germans
Cymru “Wales” and Cymry, the plural of Cymro “Welshman” are essentially the same word. The Cymru spelling for the country was invented after the sounds of u and final-syllable y fell together at the end of the Middle Welsh period. Kymry is the Middle Welsh form, for both meanings.
No difference in declension, of course, as Welsh is above all that sort of thing.
German has a lot of these (Böhmen “Bohemia; Bohemians”; Sachsen, Thüringen, Bayern, Schwaben, Preußen…), but it took me decades to even notice, because the place names have all become singular and are exclusively declined as such (Böhmens “of Bohemia”, gen. sg.; all other forms of sg. and pl. are identical to the nom. sg. anyway). The only plural placenames are those with a plural article: die Niederlande pl., vulgo Holland sg..
A couple of Kusaal placenames are formally plural, e.g.
Tɛmpaan “Tempane” (“new villages”)
Kɔlta’amis “Kultamse” (“Andira inermis trees”)
However, pronoun agreement is always ad sensum. You wouldn’t usually turn a placename into a verb subject or object; but if you did, you’d say e.g.
Anina vɛnl. “There is-beautiful.”
rather than
Li vɛnl. “It is-beautiful.” or
Ba vɛnl. “They are-beautiful.”
Treating “locative” as if it were a “gender” is very typical of Volta-Congo. Bantu languages often have several actual locative genders.
Moreover, you can’t treat toponyms, including ethnicity-based or country names, as agents (even Latin prose doesn’t like to do that):
Ghana dim da paam asʋ’ʋ-m-mɛŋ n yi Britain dim nu’usin.
“Ghana gained independence from Britain.”
(“Ghana-persons gained independence from out of Britain-persons’ hands.)
Some Dutch words have multiple plural forms with different meanings. (Helpful list
here).
bal: ballen (ball as round object) – bals (ball as formal dance)
been: beenderen (bones) – benen (legs, also bones)
blad: bladeren (leaves) – bladen (petals, sheets, magazines, …)
letter: letters (letters as orthographic units) – letteren (letters as in literature, D.Lit.)
middel: middels (waist) – middelen (means)
pad: padden (toad) – paden (path)
patroon: patroons (patron) – patronen (pattern, (firearm) cartridge)
pers: persen (press, e.g. printing press, winepress); Perzen (Persian people / cats / carpets)
portier: portiers (doorman) – portieren (car/vehicle door)
schot: schotten (dividing wall) – schoten ((firearm) shot) – Schotten (Scots)
spel: spellen (board games, game sets, matches) – spelen (games)
stuk: stuks (items) – stukken (pieces, plays (theatrical), hotties, …)
vizier: viziers (vizier) – vizieren (visor, backsight)
wortel: wortelen (carrots) – wortels (roots of any kind incl. carrots, square roots, and of all evil)
Re: Любовь vs любовь, could it perhaps be argued that the name is a zero derivation of the noun, the zero morpheme creating an invisible morpheme boundary that blocks elision of the о? (Not sure if that would actually explain anything.)
Personal names in modern Welsh usually resist initial mutation, so you say i Dafydd “to David” rather than i Ddafydd. That could be a similar integrity-preserving feature.
@gido: That’s an interesting list! From what people have posted so far, different plurals for different senses seems to be the most common manifestation cross-linguistically of the general phenomenon of morphology depending on sense. Of course though, the plural differences will come up more in some languages than others, and it looks like Dutch may have a lot of instances. That may have to do with Dutch having more than one productive way of forming plurals.
However, I also think that a few of the listed examples don’t really belong, or at least are not as interesting. I mean the ones like vizier, where the two senses are etymologically unrelated homophones.
Similarly, in German, personal names have lost their dat. and acc. forms since the 18th century. Goethe: “Ich sehe Götzen! Ich sehe Georgen!” – not anymore.
(Subverted in dia- and mesolects in which personal names take definite articles, which continue to be declined.)
Integrity-preserving? I should have thought none but an anti-Welsh bigot could have objected to initial mutation as integrity-reducing.
Well, the former is an animate noun, while the latter is not. So it could be another motivating factor.
@further examples
Polska quite obviously comes from ellipsis of polska ziemia, but it takes different case forms in comparison to feminine form of an adjective.
the zero morpheme creating an invisible morpheme boundary that blocks elision of the о
Historically the elisible vowels in Slavic languages descended from yers in a way governed by Havlík’s law: in a sequence of syllables whose vowel was a yer, the vowel in the last one disappeared, while in the preceding the yer changed to another vowel (typically /e/ or /o/ depending on the language and type of yer, in FYLOSC /a/, not sure how it works in Bulgarian which preserved ъ as full vowel), in the third from the end it disappeared again and so on. Nice example is Czech švec “shoemaker”, historically šьvьcь with three yers in a row, where only the middle one survived deletion; the genitive is ševce, historically šьvьce, here it is the first yer that survived in accordance with the law. Incidentally, there is a fairly common surname Švec whose genitive is commonly Švece, so basically the same situation as with Любов.
There is however no need to postulate a zero morpheme to explain the absence of vowel elision, since it is not regular: not every /e/ or /o/ descended from a yer and those which do not are not elided. E.g. Czech nom. pes “dog”, gen. psa (from pьsъ, pьsa) vs. nom. les “forest”, gen. lesa (from lesъ, lesa). And synchronically you can’t predict which case it is from the nominative only.
@Jerry Friedman
-or# and -orC+# are usually NORTH; -ore# is always FORCE; -orC+e# varies:
FORCE — borne divorce force forge horde
NORTH — George gorge thorpe corpse torque endorse gorse horse remorse
wikipedia “Horse–hoarse merger”
Well, the former is an animate noun, while the latter is not. So it could be another motivating factor.
Thanks, that’s a good point.
@de
In Irish, there is inital mutation, at least for Gaelic or Gaelicised names
Máire, a Mháire (voc), do Mháire (dat, “to” Máire)
Peadar, ar Pheadar (prep. “on”, “for” Peadar)
These are séimhiú. I can’t find examples of úrú for personal names, only placemames (e.g. i mBleá Cliath/ i gCorcaigh = in Dublin/Cork)
Thanks, mollymooly!
Place names may be different from personal names, like in Welsh where they are mutated too – i Gaerdydd, yng Nghaerdydd, ei Chaerdydd (is the last one actually possible?).
In contrast to that, Turkish preserves both personal and place names by putting an apostrophe between the name and suffixes, which sometimes prevents voicing (murat, murada vs. Murat, Murat’a). However this is just orthography, in speech proper names behave like ordinary nouns, if I am not mistaken.
ei Chaerdydd (is the last one actually possible?).
Sure! Assuming that the lady knows how to party:
http://www.rampantscotland.com/songs/blsongs_glasgow.htm
Alternatively, it might come from one of those ghastly articles about a Booker Prize winner: “Angharad’s view is that, though many see Cardiff as a mere pitstop on the way to the infinitely preferable city of Swansea, her Cardiff is a place of wonder …”
I’ll translate an interesting comment from the original thread that adds an extra wrench to the works. On the suggestion that there’s a proper/common noun distinction, user gul_kiev notes:
@de, Prase
I should have been more precise-the reason I didn’t have an úrú example for a proper name is that the common cases (e.g., after i “in”, after prep. ag “at”+definite article) do not naturally occur with names of persons. However, some surnames (especially Ó X, but also De X I think) have an alternate reference form with definite article (and suffix ach/each). There you can find examples, but I was not able to find them in corpora. Here is one from an early 20C book:
—
Bhí órdú fálta ag an mBrianach agus ag Seán Mandeville teacht agus a dtriail do sheasamh i dtaobh na cainnte a bhí curtha ’na leith.
—
https://corkirish.wordpress.com/wp-content/uploads/2014/11/msf-full-draft.pdf
—
This is from Mo Scéal Féin by an tAthair Peadar Ó Laoghaire, p.101 of PDF.
The “Brianach” in question was a man called Liam Ó Brian in Irish.
the common cases (e.g., after i “in”, after prep. ag “at”+definite article) do not naturally occur with names of persons
There are examples in the Historical Irish Corpus, e.g. gCaitlín or bPádraig
e.g. if Pavel is a first name, it’s Pavla, but if it’s a last name, like the Czech president, it’s Pavela
Which is kind of bizarre, since it’s “Pavla” in Czech for both first name and surname. Is this because the surname is treated as a loanword in Russian and there is a generic rule against this type of elisions in loanwords?
Slovak version is “Pavol”, “Pavla” (in Slovak ъ turned to o and ь to e), however Czech speakers tend to decline the Slovak name as “Pavola” since in Czech all yers turned to e and eliding o feels unnatural. This is considered prescriptively wrong though.
@prase
Pavla/Pavela
There are other surnames with similar pattern. It seems that those ending with -el/-ol or -eł/-oł are affected.
Kozioł – kozioł
Bargiel – bargiel
Bartel
Is this because the surname is treated as a loanword in Russian and there is a generic rule against this type of elisions in loanwords?
Maybe! It just feels unusual since it’s otherwise identical to the given name. Which creates funny results with, e.g., the Czech scholar Pavel Pavel in Russian, because you get constructions like the genitive Pavla Pavela.
@mollymooly
What a feast, thanks!
Ár gCáitlín is briefer (and has úrú) but Cáitlín ‘s againn(e) strikes me as more natural in unpolished speech.
I bPadraig– good example of the use of i to emphasise the last of two nouns in apposition!
—
“sagart mór” gan amhras a bhí i bPádraig
—
In this case, the alternative (without úrú)
—
Ba (ea) “sagart mór” gan amhras é Padraig
Or
Bhì Padraig ina “shagart mhór” gan amhras
—
is shorter but less cohesive and forceful.
The quote about Pádraig Mac Suibhne
“Ní hiongnadh linn dá réir sin go mbeadh duthchas an léighinn i bPádraig Mac Suibhne.”
is very literary, the text starts with a quote about heritage [dúchas] being stronger [treise] than study, and the quoted bit asserts that Mac Suibhne was a learned person by heritage (earlier the efforts of his grandfather and father to conserve and make available old manuscripts in Irish are described). You could also read “nature” for dúchas, but wouldn’t the text go on then about boy genius Pádraig instead of his grandfather?)
The 17C texts use the preposition ré/ria + úrú, I think this is rather archaic, modern roimh does not have úrú.
However, I also think that a few of the listed examples don’t really belong, or at least are not as interesting. I mean the ones like vizier, where the two senses are etymologically unrelated homophones.
Yes, one would have to look at the etymologies to see whether specific differences in morphology are due to differentiation for different meanings of what is originally the same word or due to two originally different words becoming homophones in the citation form.
E.g., German Bank with plural Bänke “benches” and Banken “banks” is an example of the first case, but Strauß with plural Sträuße “bouquets” and Strauße “ostriches” is an example for the second. But from the point of view of the speaker, it may be impossible to distinguish the two cases.
English has some proper nouns whose plurals are more regular than the source common noun, eg “Toronto Maple Leafs” vs “Canadian maple leaves”; or surnames like “the Salmons”, “the Bowmans”
I once knew a nursing sister once who pluralised “houseman” (a mediaeval UK term for the most junior grade of hospital doctor) as “housemans.”
My Proudfoot relations are Proudfoots rather than Proudfeet.*
I recall a discussion we once had about Blackfoots vs Blackfeet; apparently the preferred usage varies among the people(s) themselves. Though I suppose that that may be a bit different: after all, there’s nothing illogical about a singular Proudfeet, given the usual human complement.**
* Reminds me that when Michael Foot visited Romania many years ago, I’m told that the Romanian media scrupulously pronounced his surname as [fo:t].
** Kusaal does this with some bahuvrihis, e.g. bitʋbli’id “deaf child” (singular), which breaks down as bi- “child”, tʋb- “ear” and the plural deverbal adjective form li’id “blocked up” (li’ “block up.”)
Good one, mollymooly. I wonder why I thought of the ones where the names have irregular plurals, but not the ones where the words that became names had irregular plurals.
@DE:My Proudfoot relations are Proudfoots rather than Proudfeet
I had imagined you were taller.
Perhaps I may quote a notable translation fail, for those who read Spanish.
Here in New Mexico, where surnames not ending in -s or -z get regular plurals (which would shock the entire rest of the Spanish-speaking world if they knew about it), you might actually be able to translate that with “Ganapiés” and “Ganapieses”.
I had imagined you were taller
Not according to Elon Musk.
I’ve seen field mouses from a native speaker. Mouse is on the list of nouns with irregular plurals, field mouse is not…
Sabertooths for saber-toothed cats is universal, under the excuse that it doesn’t refer to the actual teeth. (Would never fly in, say, German.)
—¡Ganapié! —gritó un viejo hobbit
It’s…
—¡Ganapiés! —gritó un viejo hobbit
…in my ©2016 copy. (Still “Ganapié, repitió Bilbo”)
IIRC, the Spanish dubbing of the Ralph Bakshi versión I loved as a kid had Bilbo saying Ganapatas and Odo amending Ganapiés
It turns Bilbo’s mistake into what can only be an intentional jibe, but it worked as humour, at least for an audience of children
JF’s alternative is better, though; pieses is common if nonstandard in lots of dialects, and I’d be surprised if that wasn’t the case already at the time those translations were made
@mollymooly: That works a lot better. You can see the version I quoted at https://www.facebook.com/tolkienysuliteratura/posts/mis-queridos-bols%C3%B3n-y-boffin-comenz%C3%B3-nuevamente-y-mis-queridos-tuk-y-bolger-y-br/1189687954846805/
and if I knew someone with an unlicensed copy of El señor de los anillos on their computer, it would also have an unchanged “Ganapié”.
@Alon: I see you prefer the spelling -ses for that non-standard plural suffix, and it makes more sense to me than -ces, but I think I’ve seen -ces more often (in my limited reading). Do you or does anyone else have any thoughts on why one is better?
@JF: I think it has to be <s> because the peninsular speakers who use these forms have /s/ rather than /θ/
@Alon Lischinsky, @Jerry Friedman:
The RAE (Nueva gramática de la lengua española § 3.2z) only says the following:
I don’t know if vulgar Spaniards use the non-standard plural, but of course around here Alon is right that spelling and pronunciation are one to one.
Here’s Fundéu castigating Americans on a related point: