TL,DR: We don’t say Mercedeses because that is not how we normally speak English today.
- How many Saint Diomedes have there been?
- How many Antilles have you visited?
- How many Portuguese are on board?
- How many Mercedes did Mick Jagger own?
- How much does this Mercedes cost?
- How many Saint *Diomedeses have there been?
- How many *Antilleses have you visited?
- How many *Portugueses are on board?
- How many *Mercedeses did Mick Jagger own?
- How much does this *Mercede cost?
That is my opinion on what sounds right and wrong to me as a native speaker, following only the internalized ruleset of my own ear. Reasoning and documentation follow.
The singular Mercedes ≣ the plural Mercedes
First off, any discussion about what is or isn’t done with the word Mercedes in German or in Spanish doesn’t matter. Here’s why:
- Because we are speaking English, not either of those other two languages.
- Because this is a brand name, a marque of automobile, it is more likely to follow its own private little usage customs than a regular word is.
- Because of deeper phonologic patterns that we should not be deaf to.
- Because of the massive preponderance of common usage that cannot be ignored.
More detailed elaborations will follow, but let me provide right up front
one prominent reference supporting Mercedes as the plural.
According to Wiktionary’s entry for Mercedes, the word is a proper noun whose plural is also Mercedes. They write:
Spanish Mercedes. The car was named for Mercedes Jellinek, the daughter of Austrian businessman Emil Jellinek who ordered 36 cars from Gottlieb Daimler.
Mercedes (plural Mercedes)
- A female given name occasionally borrowed from Spanish.
- (trademark) Short form of Mercedes-Benz.
There you have it, and in print no less. So if you are convinced, or merely impatient, then that answers that and you can all go home now.
Very well, then.
It is true that the Wiktionary editors do not explain why its plural form happens to be the same as the singular version, nor do they provide any references for their position. (They never do, though. It’s just how Wiktionary is: no references, ever.)
So to back up what they have written, I below explain in some small measure of detail why I think they said what they said and why I agree with them, presenting different arguments and counterarguments, plus of course the requisite N-gram that’s such the rage in these parts.
You might want to get a cup of coffee first, though.
What’s going on here is that broadly speaking, words that end in unstressed
/i:z/ in English do not normally gain anything when inflected for the plural, the possessive, or both. That’s also why a plural possessive only gets one marker, not two.
Both inflections add
/əz/ to the word, but dups are suppressed. The unstressed final
/i:z/ also suppresses the further addition of an additional
/əz/, no matter whether it is for making plurals or for making possessives, or both.
- two species
- this species’ name
- both species’ names
- the series’ final episode
- both series are great
- both series’ finales sucked
- this chassis is new (final s is silent there in the singular, so pronounced
- both chassis are new (in the plural, the final s becomes pronounced, so it’s either of
- various Saint Diomedes (compare: various Saint Johns and Saint Annes)
- many Hercules are recorded
- Hercules’ labors
- Socrates’ choice
- Achilles’ rage
- Ulysses’ son Telemachus
- Eratosthenes’ sieve
- this Mercedes is great
- this must be Mercedes’ best car
- both Mercedes are great
- both Mercedes’ batteries are missing
You might be tempted to think that words like trapeze and chemise are “exceptions”, but they aren’t because their
/i:z/ there is stressed. This law of suppression of inflectional
/əz/ only operates when there is a terminal unstressed
/i:z/, not a stressed one.
For some speakers a terminal
/əz/ also suppresses the addition of another inflectional one. For other speakers, it does not — which is what leads to dialect spellings of things like the farmerses fields or tasty fisheses (as Gollum, ever the rustic in speech habits, might say).
However, for all standard speakers it will suppress an additional
/əz/ just to make it possessive if already plural, and vice versa. So all three of church’s and churches and churches’ are pronounced the same, and there’s no such thing as the *churcheses’ taxes, no matter whether in speech or in writing.
“Are there any Chineses here?”
“Is there a Chinee in the house?”
English does not have many count nouns that end in
/i:z/ in the singular, the way species, series, and chassis all do. And when we do have them, they do not gain an extra ‑es to make them plural for the phonologic reasons outlined above.
There has, however, been some historical variability with regard to that, particularly during the 17th century. Probably the largest class of such
/i:z/ words in English are those used for languages and people, words like Chinese, Japanese, Portuguese, Siamese, Pekinese. Once upon a time, we actually did sometimes add an extra
/əz/ to those to make them plural, but that is now considered obsolete.
As Janus has pointed out in comments, one can analyse this set a bit differently from how one analyses words like species, series, and Mercedes. The difference is that the final
/i:z/ syllable in the language/people words is stressed, which puts them more in the trapeze and chemise class instead.
However, they have all now fallen into the same pattern as the first set, and all those nominalized adjectives for the people of that place are today “invariant” in form, meaning that they can be used interchangeably as either a singular or a plural without changing the world itself.
Janus has also reminds us in comment that while we no longer say several Portugueses to refer to several people from Portugal, we can still use Portugueses to distinguish multiple varieties of the language, a much less common occurrence. It’s like saying there are multiple Englishes or multiple Frenches: it today applies only to languages not persons, and is not commonly heard. A realistic place you might say it would be that “the Portugueses of Europe and Brazil differ in various ways”. But you can’t ever do that with Mercedes because its
/i:z/ is no more stressed than the one in species.
Clarifying another point made in comments, these “<CAPITAL LETTER> + ‑ese” words like Chinese and Portuguese are not plural-only forms the way people, pants, and scissors normally are. It just means that the singular and plural look the same, not that there is no such thing as a singular the way occurs with the just-mentioned pluralia tantum.
So you can have one Pekinese is, many Pekinese are, in the same way that you can have one chassis is, many chassis are or one sheep is, many sheep are. Your Mercedes works the same way: one Mercedes is, many Mercedes are.
Here are a few specifically plural examples of “<CAPITAL LETTER> + ‑ese” words from the OED’s citations for each:
- 1697 Dampier Voy. I. xv. 406 ― The Chinese in general are tall.
- 1783 Watson Philip III (1839) 133 ― The affairs of the Portuguese in India were more than ever neglected by the government at home.
- 1797 Encycl. Brit. (ed. 3) XVII. 449/1 ― The Siamese prepare the land for tillage as soon as the earth is sufficiently moistened by the floods.
- 1797 Encycl. Brit. (ed. 3) X. 492/2 ― The Maltese still continued to behave with their usual valour against the Turks.
- 1797 Encycl. Brit. IV. 310/2 ― The Ceylonese make use of boats hollowed out of the trunks of trees.
- 1861 Chambers’s Encycl. II. 575/2 ― The Cantonese are notorious for their turbulence and hatred of foreigners.
- 1894 Parry Stud. Gt. Composers, Beethoven 166 ― His behaviour was not of the kind affected by polite Viennese.
- 1906 Field 20 Oct. 663/2 ― Pekingese were forward in strong numbers, the best dog weighing [etc.].
- 1920 Glasgow Herald 5 Apr. 6 ― The Lebanese··have··dissociated themselves entirely from the action of the Syrian Congress.
- 1969 J. Wood Three Blind Mice iii. 42 ― Taking on two belligerent Faroese.
During the 1600s, however, there was a tendency among some writers to use plural forms like Chineses and Portugueses for several of each, but that is many centuries in our past and nobody does that any longer. The OED writes of Portuguese:
- A native of Portugal. The plural Portugueses (‑guezes) was used during 17th c.: since it became obs. Portuguese has been sing.
and pl.; in modern times a sing. Portug(u)ee has arisen: see Portugee. Cf. Chinese>Chinese, etc.
With this example of the obsolete plural:
- 1694 W. Wotton Anc. & Mod. Learn. (1697) 269 ― The Portuguezes, who first made daring Voyages, by the Help of the Compass, into the Southern and South-Eastern Seas.
And with this definition for the “spurious” singular, including citation:
Portuguee, Portagee, Portugee. Repr. a spurious ‘singular’ form of Portuguese a. and sb., this being regarded as a plural.
- 1878 Besant & Rice Celia’s Arb. xxviii, ― A Portugee, as every sailor knows, is a Portugee by birth.
The same is true of Chinese:
B 1. a. A native of China. The plural Chineses was in regular use during 17th c.: since it became obs. Chinese has been sing. and pl.; in modern times a sing. Chinee has arisen in vulgar use in U.S. (So sailors say Maltee, Portuguee.)
With this example of an obsolete plural Chineses:
- 1667 Milton Paradise Lost iii. 438 ― Sericana, where Chineses drive With Sails and Wind, thir canie Waggons light.
And this example of singular Chinee, derived the same way as Portugee was:
- 1871 Bret Harte That Heathen Chinee, ― The heathen Chinee is peculiar.
Unlike “foreign” loanwords that may be confusing as to their number, those “<CAPITAL LETTER> + ‑ese” words are all “real” English words. They still sometimes get reänalysed as being in the plural, leading to “spurious” back-derived singulars like Chinee and Portugee.
This is the same sort of reänalysis that gave us a pea as a back-formation from the earlier pease
/pi:z/ (from Latin pisum), which could be either singular or plural. So folks saying pease as a plural created a new singular, pea.
It is so natural to think of an
/i:z/ word as a plural that confusion often arises, and “spurious” forms may appear.
For example, the Greek word litotes, a word perhaps not so well known to the general public but quite common on ELU, is singular but has sometimes been mistaken for a plural, as in this OED citation:
- 1645 Rutherford Tryal & Tri. Faith xv. 116, ― Ps. 23. 4 Yea though I walk [etc.]; its a Litote, I will believe good: its a cold and a dark shadow to walke at deaths right side.
This same confusion often happens with other imported
/i:z/ singulars. For example, the botanical term for a plant stalk is a stipes, with the original Latin plural of some stipites being observed by Latin-trained botanists.
But stipes was reänalysed as a plural of a putative stipe and so now stipe > stipes is commonly seen as a “naturalized” English (or even French) version of Latin stipes > stipites. The OED gives all these examples:
- 1796 Withering Brit. Plants (ed. 3) I. 84 ― Stipes, a pillar, or pedicle. Also the stem of some kind of Fungi.
- 1797 Encycl. Brit. (ed. 3) XVII. 597 ― The stipites or younger branches are directed for use, and may be employed either fresh or dried.
- 1847 Henfrey Outl. Bot. 114 ― When the indusium is torn by the expansion of the pileus and the elongation of the stipes or stem.
- 1864 T. Moore Brit. Ferns 10 ― The fronds of Ferns consist of two parts-the leafy portion; and the stalk, which latter is called the stipes.
- 1871 W. A. Leighton Lichen-flora 41 ― Stipites and capitula cinereo-suffused.
- 1861 H. Macmillan Footn. Page Nat. 214 ― The tubercle rapidly increases, until at last it produces from its interior, a long, thick, fleshy stem or stipe, surmounted by a pileus.
Nowadays the “regular” stipe > stipes version has come to be more common than the original but “irregular” stipes > stipites version.
Um, so what?
The moral of the story is that all
/i:z/ singulars in English are sufficiently deceptive-sounding that strange things do, can, and often enough have happened to them. People both over- and under-pluralize them quite commonly, and how (and if) it finally settles out is never a complete given.
Documentation for virtual non-use of *Mercedeses
As for Mercedes or Mercedeses, while Google books has many instances of both Mercedes and some Mercedes, it has absolutely no instances of those use with in would-be *Mercedeses form.
If you look for two Mercedes versus two Mercedeses, you do find a few. You also discover that *“two Mercedes outnumbers the latter by a factor of twenty to one:
A 20:1 majority is such a tremendous landslide in the popularity metric that most speakers would consider the minority use spurious. That is, they would think it is incorrect or wrong or awkward or silly sounding.
Here are more recent published examples, mostly from newspapers:
- The two Mercedes were predictably in a different league to their rivals and recorded their second one-two finish in eight days after Hamilton's victory last weekend in Malaysia. [Associated Press]
- Seven Ferraris, two Mercedes and a Ford: the world's priciest vintage cars [Hindustan Times]
- Sarasota Classic Car Museum Photo: Two Mercedes Owned by John Lennon, Sarasota Classic Car Museum [Trip Advisor]
It is no surprise that Wiktionary documents the plural form of Mercedes as being the same as the singular Mercedes, because that is what people are using.
Interestingly, one of the hits in Google Books for the Mercedeses version is actually spelled Mercédeses with an acute accent, as this is the proper plural for two girls with that Spanish given name (see Appendix I). This occurs in The Commanding Heights: The Battle for the World Economy
by Daniel Yergin and Joseph Stanislaw. Using the Google Books hit for the 2002 edition, you can see that it reads:
At street level, youths—Malay, Chinese, and Indian—clad in jeans and T-shirts zoomed on motorbikes through streets jammed with Japanese cars, Mercédeses, and the ubiquitous Proton, Malaysia’s national car.
This example is an especially interesting one for several reasons. It is comparatively recent, occurring since the turn of the millennium. And it uses a very specifically Spanish way of writing that word, complete with the mandatory acute accent it requires in that language. This signals that they were not treating it as an assimilated English word at all, but rather as a Spanish import that carried its own inflectional rules with it, ones requiring “exotic” diacritical embellishments.
We don’t say Mercedeses because it is not how we normally speak in English. It sounds as wrong to most of us as saying Chineses for several people now does.
Just because you can find random counterexamples on the Internet proves nothing whatsoever.* Not only can you find any nonsense you please on the Internet, you can also find any nonsense that doesn’t please you, too.**
* Otherwise teh would be an acceptable spelling of the — and it isn’t.
** As Charlie Stross showed in his very-near-future novel, Rule 34. :)
There is no real need to read any further, but I have a few notes I was making as I was drawing this up, so I might as well set them down here in case it helps illuminate the paths I trod in coming to my own conclusion.
Appendix I: Spanish plurals
Something was said in comments that I wanted to clear up. It was said how in the word Mercedes is plural in Spanish.
Well, yes and no.
It is true that more than una merced in Spanish would be unas mercedes using the standard rule of adding ‑es to singular common nouns ending in a consonant to make them into a plural.
But that is only operative as a common noun, not a proper one.
And that is not what the capitalized word Mercedes is at all: it’s a proper noun, and proper nouns have their own pluralization rules in Spanish, just like they often do in English.
It turns out that the formal rules for making Spanish plurals are trickier than most people think. Please see section 2.8 about proper nouns in the referenced Reglas de formación del plural from the Real Academia Española, which although nowhere near as awesome as our own Royal Academy, is more suitable for this situation.
For Mercedes, you really have to consider three cases:
- Section 2.8a: When Mercedes is used as a given name — “un nombre de pila” — then it would take the regular plural of Mercédeses, and you might end up knowing varias Mercédeses in Spanish just like you might know various Janices (or even Januses) in English.
- Section 2.8b: However, when Mercedes is used as a surname or family name — “un appellido” — like Smith in English, then although in English we talk about “the Smiths down the street” with a plural ‑s suffix, they in Spanish would leave the surname untouched but still use plural concord for the family of los Mercedes que viven por acá. (There is some variation though if you are just talking about different Mercedes families.)
Section 2.8d: When Mercedes is used as a brand name — “una marca comercial” — then it is normally left as is since it already ends in ‑s. The RAE provides the following illustrative example of an invariant brand plural:
- Los nuevos Corte Inglés de la ciudad son muy grandes.
Notice how it is not “los nuevos Cortes Ingleses”. They just don’t do that with brand names. And so for the cars it would also be los nuevos Mercedes, too. (Yes, the girls are feminine but the cars are masculine.)
Appendix II: Style-Guide Disagreements
You can find people who disagree with Wiktionary saying that the plural of Mercedes in English is still Mercedes. This proves nothing by itself, since you can always find people who disagree with anything. People are like that.
Now sometimes this disagreement is based on actual reasoning, but most of the time it is not. And even when based on reasoning, that reasoning may be faulty, just as it was when people who should have known better argued, often at great length, that “infinitives should not be split in English”, or that “you must not end a sentence with a preposition”, or that “you must never use between for more than two items”.
In all those cases, the person making that argument thought that they were right and that anybody and everybody who did it otherwise was wrong — no matter how many native speakers didn’t agree with them. They make no attempt to describe what people actually do, or recognize that different people do different things and that this is natural, normal, and perfectly fine. Look at the wars over data is/are for a more recent battle royale by language mavens who think they know better than the people using the language. Also check out the “n-grams squishiness” section in that answer.
Worst of all, these usage guides became confused with actual rules of grammar — which they were not. Real rules of grammar are things like verbs agreeing with their subjects, or using this/that with singular nouns but these/those with plural nouns.
So what happens is you have people saying what they think you should do, but this advice is taken as some sort of Eleventh Commandment dispensed by the Royal Academy of the English Language. That’s what happened with all those cases I just mentioned above, and we’ve spent centuries trying to recover from all that damage.
If you look hard enough, you will see that same thing happening here too. Possibly the most damaging disagreeing voice is that of Bryan Garner in his eponymous Garner’s Modern English Usage.
Even though Garner is no Fowler — and even if he were, it would change nothing — there are those who treat his words as enforceable laws dispensed on golden tablets by the Royal Academy of the English Language.
And so it must be noted how on page 533 of his book, Garner makes the (weak) claim that the plural of Mercedes has to be Mercedeses. It is weak for several reasons, and he himself even recognizes that this is not what people actually say. But he nevertheless pretends that they “must” be guilty of “[flouting] the age-old rule for pluralizing names.”
In my opinion, this is an area Garner has not thought through. Furthermore, it sounds just as dismissive of others’ actual usage as Goold Brown was about between.
Garner fails to recognize that imported words like series, species, chassis, Socrates, or Ulysses are all operating under different sound-rules in English, let alone does he show any understanding of why this might be. So he doesn’t notice that Mercedes is another one of that same class, and provided it counts as an assimilated English word, that it should work just like they all do.
I recommend that you ignore Garner’s unsound advice in this, even if you should have two different Socrates in the room with you. Nobody speaking today’s English should be saying Socrateses or Pekingeses or Portugueses or specieses or chassises or serieses — or Mercedeses. Blind obedience to a simple-minded “rule” that ignores both actual usage and deeper morphological patterns is just so much poppycock and fiddlesticks.
It’s like how you are never going to get universal agreement on pronunciation across the entire Anglosphere. Even seemingly simple names like Oregon, Paris, and Missouri are constantly (and uselessly) debated. Proper names borrowed from elsewhere into English are especially troublesome: just look at the controversies surrounding the “right” way that one should pronounce Hermès, Nike, Nikon, Audi, Volkswagen, and Porche when speaking English.
Feel free to add plural Mercedes to that long list of controversial usages if it pleases you to do so. As for me, I think it fits in just fine with existing models, and I believe I have demonstrated why this is so.
Your goal is to be understood. Insisting on a simplistic and facile non-rule “rule” despite actual usage is just going to get us peopleses looking at youse funny like.