1

This is Konrad.
He has a dog.
Hence, it's Konrad's dog.

This is someone else.
He has a cat.

  1. Hence it's someone else's cat.
  2. Hence it's someone's else cat.
  3. Hence it's someones else cat.
  4. Hence it's someone elses cat.

I know that I can say "it's a cat of someone else" or "it's a cat belonging to someone else" but that dodges the issue, not answering the question.

  • 1
    In addition to the answers already given, it is also not correct to say, “This is a Konrad” (unless a Konrad is some species of animal I’ve never heard of); you just say, “This is Konrad”. And you cannot say, “it’s a cat of someone else” unless the cat has actually been made from the chopped up pieces of someone else’s body (which is somewhat unlikely in most non-psychopathic circumstances). The ‘of’ construction is not used for actual possession/ownership with animate beings. – Janus Bahs Jacquet Jul 26 '13 at 21:44
  • @JanusBahsJacquet What?! In Sweden we chop up corpses and build cats all the time. :) Seriously, though - good point. I corrected "a" for Konrad. It was "a house" from the beginning. – Konrad Viltersten Jul 26 '13 at 21:48
  • To paraphrase for Swedish, whose genitival clitic -s functions in the same way as the English one: ‘someone else’ = en annan. ‘Someone’s else cat’ = ens annan katt (clearly wrong). It must be ‘someone else’s cat’ = en annans katt. – Janus Bahs Jacquet Jul 26 '13 at 21:48
  • 1
    @Carlo_R., no, not at all. Swedish (like English) is a language belonging to the Germanic branch of the Indo-European family; Arabic is a Semitic language. En annan is in no way Arab. It is, in fact, an exact cognate to English ‘another’ (except ‘other’ is an old nominative, where Swedish annan is an old accusative). The fact that something looks or sounds like something in a different language in no way implies that the languages are related (compare how Greek θεός theó-s ‘God’ is very similar to teō-tl in Nāhuatl, even though they are obviously completely unrelated languages). – Janus Bahs Jacquet Jul 26 '13 at 23:10
3

According to The Cambridge Guide of English Usage, so well established are phrases like "someone else", "anyone else", "what else" and "who else" that else can take the possessive form quite easily.

In the light of the above reference the possessive of "someone else" is "someone else's".

It is worth noting that "This usage was once frowned on by those who insisted that else was an adverb and so could be made possessive", and this presumably justifies your doubts.

  • 1
    Carlo, this doesn’t actually have anything to do with else. It’s that the ’s attaches to the entire NP in English, not just to the N. – tchrist Jul 26 '13 at 21:40
  • Carlo, someone else is an NP. – tchrist Jul 26 '13 at 21:53
4

The Modern English possessive suffix -'s is not a case any longer. Cases inflect nouns, but the -'s attaches to the end of noun phrases, rather than to their head nouns.

Technically, an affix that attaches to a syntactic construction instead of to a word of a particular type is called a clitic (a clitic can be either an enclitic or a postclitic, just like an affix can be a prefix or a suffix; respectively).

So, as an NP postclitic, -'s attaches to the end of the last word of the noun phrase whose head noun is the possessor. Thus, the Queen of England’s first great-grandson or the guy at the door’s keys, as tchrist points out -- it's not England's g'grandson, it's the Queen's; and they're not the door's keys, they're the guy's keys.

Thus, also, Somebody Else's Problem.
Indeed, as HHGttG makes clear, the SEP field is very advanced technology.

  • 2
    Here, John, I bet you’ll appreciate this. For the record, clitics can also be mesoclitics. In Portuguese, unlike in (Modern) Spanish, the clitic pronouns fall between the stem and the inflection’s ending in the future and conditional. Yeah, it’s weird. It’s because of how modern Romance synthesized new tenses out of postpositioned auxiliaries after Latin infinitives. See here and this scholarly article. – tchrist Jul 26 '13 at 22:05
  • Yes, a sandwich of fossilized cliticization in all those cases. Kinda like shouldna or could'ven't, both of which I've observed. – John Lawler Jul 27 '13 at 0:31
2

The answer is obviously that it’s someone else’s cat, just like it’s the Queen of England’s first great-grandson or the guy at the door’s keys.

I honestly don’t even understand how the question could ever come up.

  • 2
    NP = noun phrase. Postclitic is a clitic (a small pseudo-word that can only appear if attached to a proper word) that is added to the end of a word (rather than before its beginning, in which case it’s a preclitic). – Janus Bahs Jacquet Jul 26 '13 at 21:45
  • 2
    Also, a correction to my comment above: the term is of course proclitic (not †preclitic). – Janus Bahs Jacquet Jul 26 '13 at 21:54
  • 1
    This is not a question that would come up with a native speaker. They might ask why it isn't *someone's else problem, but they'd know what it should be. A non-native speaker who knew an inflected language might believe what was in their textbook, however. – John Lawler Jul 26 '13 at 21:55
  • 1
    @tchrist, they are told it means the possessive, yes, but they know instinctively where the clitic goes because they already speak the language and use the clitic. They do not learn, before actively using it, that ’s is a genitive case. It is quite understandable that someone who is learning English through grammar and is told that ’s is a case ending would expect ’s to then act as a case ending, rather than as a clitic. – Janus Bahs Jacquet Jul 26 '13 at 22:00
  • 1
    @JanusBahsJacquet I doubt that grade-school teachers are allowed to even say the word clitic in class. :) – tchrist Jul 26 '13 at 22:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.