I'm trying to determine the part of speech in the following example:

German: Mario Götze ist einer der besten Fußballspieler der Welt. (partitive genitive)

English: Mario Götze is one of the best soccer players in the world.

Thanks for your help!

  • 2
    Looks like it's a pronoun, in the plural it changes to "some": They are some of the best players. In Russian also the word for "one" is used in such a construction, and in the plural it has a plural ending, which is a definite sign it's not a numeral.
    – Yellow Sky
    Jul 30 '14 at 13:16
  • 1
    I agree. In English, one is an indefinite pronoun (not completely assimilated to a pronominal paradigm yet, though -- the genitive is one's). It also appears in the plural with a regular inflection -- the ones I told you about. These particular sentences, however, are much more complex than they look -- comparatives with phrasal determiners are not simple syntactically.
    – jlawler
    Jul 30 '14 at 14:52

Interesting question!

Tim Osborne has produced a lengthy answer. A response to his answer requires more space than is available in comments. I have therefore modified my own answer in order to produce a response. The response follows my answer.


German and English are too different, with German exhibiting case/genus morphology via its inflectional suffixes, that is simply absent in English. Hence determining which part of speech one is dealing with, may yield conflicting cross-linguistic results. The answer below concerns the German example.

I think German einer is still an article. It appears with the nominative inflection -er, which is blocked whenever a noun appears together with it.

  (1)   Da ist ein Mann.
  (2)   Da ist einer.
  (3) * Da ist ein.  

The nominative appears, of course, because the entire post-verbal construction is the predicate noun, which must take nominative case.
The phenomenon exists not only with the indefinite article ein, but also with the negative article kein:

  (4)   Kein Spieler fehlt.  
  (5)   Keiner fehlt.
  (6) * Kein fehlt.

Obviously, the indefinite article ein, and the negative article kein belong to the same subclass, namely that requiring the mixed declension.
The interrogative article welch, however, belongs to the strong declension:

  (7)   Welcher Spieler ist schon ausgewechselt?
  (8)   Welcher ist schon ausgewechselt?
  (9) * Welch ist schon ausgewechselt.  

The examples (1-9) are all in the nominative masculine because the example in the question is, too. Other inflections are, of course, possible, depending on genus and case.
The property the examples (2), (5), and (8) share with the example in the question is that a noun fails to appear with the article. As a result, I would view the syntactic heads of the partitive genitive expressions still as articles.
The plural of these construction is, by the way, also an article, namely einige. However, other articles are possible, too, such as viele, or wenige, or constructions such as ein paar. These expressions are articles.


Tim Osborne interprets my answer (above) as invoking noun ellipsis. Doing so is his prerogative, but that doesn't make it true. In fact, I do not claim that such a form of ellipsis is involved, rather I state that "a noun fails to appear with" forms such as einer. The attempt to construe that as the advocacy of noun ellipsis is far-fetched.

Tim Osborne seems to view his example (3b) as evidence for his position. While (3b) is indeed ungrammatical, there is a grammatical alternative that Tim Osborne fails to mention:

 (10) Sie sind welche (*Fußballspieler). 

(10) now has an article, namely the one that must appear in the plural of indefinite nouns and of uncountable nouns. To construe Tim Osborne's (3b) as something I would be forced to uphold is a strawman-argument.
Tim Osborne's examples (4) are irrelevant because they show VP-ellipsis in English. These data do not contribute to an understanding of the German data.

There are two counter-arguments against the pronoun hypothesis:

1.If einer is a pronoun, rather than an inflected article, then two scenarios are possible.

A. The pronoun is morphologically exocentric. That means that the word/part of speech class of einer cannot be derived from the combination of ein, an article, and -er, or other inflectional affixes. Since -er is an inflectional suffix, it cannot cause derivation. We would have to assume

 (11) Art+Infl = P

Thus we are forced into the assumption of exocentricity.

B. One could assume the pronominal variations are morphologically simple in form, but complex in meaning. That means one could assume that the word/part of speech class is simply P, rather than X-Infl, but that assumption would not hold for semantics. One would then have to assume to different sets of these expressions, for instance einem as an article and as a pronoun, marking non-feminine dative singular. Thereby one would violate Occam's Razor: Pluralitas non est ponenda sine necessitate, i.e. 'A plurality is not to be posited without necessity'.
It seems more parsimonous to assume that articles enact pronominal functions whenever they appear in the absence of a superordinated noun, regardless of this absence being caused by noun ellipsis or not.

2.The adverb so can precede German articles, but it cannot precede pronouns.

 (12)   So einer gefällt mir.
        lit. 'I like such one.'
 (13) * So er gefällt mir.
        lit. 'I like such him.'

Examples such as (12,13) could be produced in much greater number, for every case and every genus. The fact that the adverb so can precede einer, but not er would require an explanation. If einer continues to be an article in (12), then such an explanation is easily accomplished.

My proposal is that words such as einer in the partitive constructions continue to be articles. Nouns are not elided, they are entirely absent.
Tim Osborne's response is built on a strawman-argument, namely that I propose noun ellipsis. I did not.
Tim Osborne's answer incurs either the problem of exocentric morphological structure, or a violation against Occam's Razor.
It also fails to deal with the adverb so, which can precede articles, but not pronouns.

  • The question can be understood in terms of N-ellipsis (=NP-ellipsis). Does a sentence such as "Er hat einen" involve N-ellipis or not? If it does, then "einen" is an indefinite article. If it does not, then "einen" is an indefinite pronoun. I prefer the latter stance, for reasons I will elaborate on later, I hope. In other words, I disagree with your answer. Jul 30 '14 at 22:17
  • I think your response has misinterpreted some of my points, especially concerning example (3b). In a nutshell, your reasoning would force us to assume that, for instance, "Lachen" 'laugh' is a verb in the following sentence: "Das Lachen war laut" 'The laughing was loud'. Morphologically, "Lachen" looks like a verb, but it is clearly functioning as a noun. Your position sees morphology trumping syntax, whereas my position sees syntax trumping morphology (when the goal is to assign given tokens of words to syntactic categories). Aug 3 '14 at 23:17
  • @TimOsborne Lachen is a (derived) noun because the infinitive functions as a derivational suffix; it changes verbs into nouns with neuter genus. Syntax and morphology need to be in sync. Aug 4 '14 at 6:15
  • OK, I agree. But why doesn't the same reasoning work for "-er" in "einer" (*einer Fussballspieler vs. einer)? Aug 4 '14 at 10:21
  • @TimOsborne I didn't introduce that example, you did. But here's the answer, anyway: -er cannot be a derivational suffix that produced derived pronominals because then the attribute adjective in ein guter Spieler would have to be viewed as a pronominal. The ins-and-outs of inflection within the German NP are way too complex to be treated in a thread. If you believe this approach is worth continuing I suggest posting a question. Aug 4 '14 at 11:29

I agree with Thomas that the situation in English and in German is not the same. I do not, however, agree that “einer” (nom. sing. m.) is an article. The indefinite article in the nom. sing. m. is “ein”. “Einer” is primarily a cardinal number. “Einer der besten Fußballspieler” is parallel to “Zwei der besten Fußballspieler”.

In both languages “einer/one” functions in certain contexts as a pronoun (“einer ist gekommen” in the sense “someone came”), but English has gone further than German in this direction. Thus in English you say “the big ones and the little ones”, but in German you cannot say this; you need to say “die großen und die kleinen”.

  • I disagree. The cardinal number '1' in German is eins. Further, there isn't just einer, but a whole system: einer, eines, einem, einen just for masculine. To this one has to add feminine and neuter forms. kein and welch also have plural forms. Jul 31 '14 at 14:15
  • "Wieviele Menschen stehen im Zimmer?" "Nur einer."
    – fdb
    Jul 31 '14 at 14:18
  • Or equally possible "Nur eine Frau". I don't deny that German indefinite articles also appear in contexts in which they function like numbers, or like pronouns. But that's just one of the roles they can take. That doesn't make indefinite articles numbers or pronouns. In light of the entire discussion, however, the proposition that ein(er) is a cardinal number doesn't really point toward a solution because we also find kein(er) and welch(er). An answer must take those articles into account, too. Jul 31 '14 at 17:30
  • Etymologically “eins, einer etc.” and “one” are numbers, with absolutely clear-cut cognates in all Indo-European languages. In German, as in Sanskrit, Greek, Latin etc. the word for “one” is fully inflected for case and gender. So, the fact that “einer” in inflectable says nothing about what part of speech it is. The use of “ein, eine etc” as an indefinite article is secondary, as is the use of “einer” and “one” as pronouns.
    – fdb
    Jul 31 '14 at 18:47
  • I suggest you include what you wrote in your comments in a revision of your answer. That would serve to make it a coherent answer. At the moment it looks more like a comment. Aug 1 '14 at 9:19

The question can be explored in terms of N-ellipsis (noun ellipsis, also known as NP-ellipsis). N-ellipsis is illustrated in the following b-sentences:

 1a. Er ist ein Fussballspieler. 'He is a soccer player.'
 1b. Er ist einer. 'He is one.'

 2a. Wir kennen einen Fussballspieler. 'We know a soccer player.'
 2b. Wir kennen einen. 'We know one.'

Based upon the a-sentences, one can argue that the b-sentences involve ellipsis; one assumes that the noun Fussballspieler has been elided (=omitted). The ellipsis account would assume that a null node is present in the syntax, occupying the position of the elided noun. If ellipsis has occurred in this manner, then the forms einer and einen are indeed articles as Thomas Gross' answer suggests, since there is in fact a covert (=elided) instance of Fussballspieler present. The primary syntactic trait of articles (definite and indefinite) is that they introduce nouns. The forms einer and einen would be classified as instances of the indefinite article by virtue of the fact that they are in actuality introducing a noun, albeit a covert noun.

The alternative analysis of such cases argues that ellipsis has in fact not occurred. It sees the forms einer and einen in the b-sentences as indefinite pronouns. The primary trait of pronouns is that they occupy the position of a noun, referring to an earlier (or later) instance of the noun in context that they replace.

In my view, there are solid emperical observations demonstrating that the latter analysis should be preferred. In other words, the einer in the example sentence in the question, i.e.

 Mario Götze ist einer der besten Fußballspieler der Welt.
 'Mario Goetze is one of the best soccer players in the world.'

is an indefinite pronoun; it is not the indefinite article. I therefore disagree with Thomas Gross' answer and mostly support fdb's comment.

The first observation supporting the analysis of einer and einen in the b-sentences above as indefinite pronouns is the systematic variation that occurs in inflection with such words. The variation is visible across sentences 1a (ein) and 1b (einer). The syntax is marking when the article/pronoun functions as an article and when it functions as a pronoun by way of the inflectional endings - vs. -er.

A second observation supporting the anslysis of einer and einen in the b-sentences above as indefinite pronouns is the fact that if such cases truly involved ellipsis of a noun, one might expect ellipsis to be possible in the absence of the article altogether. But such attempts at ellipsis fail, e.g.

 3a. Sie sind Fussballspieler. 'They are soccer players.'
 3b. *Sie sind. 'They are.'

If ellipsis of nouns were in general possible, one would expect 3b to be acceptable. In other words, the ellipsis analysis requires that one account for the fact that ellipsis in such cases requires the presence of an overt word (e.g. the indefinite article) introducing the ellipsis.

A third observation is particularly telling. This observation concerns the partitive genitive example in the question. A primary trait of ellipsis in general is that it occurs optionally. This point is illustrated with an instance of VP-ellipsis:

 4a. He will call, and she will call, too.
 4b. He will call, and she will, too.

Both of these sentences are perfectly acceptable. Whether or not one elides call is up to the speaker. Given this observation about ellipsis in general, examples of the partitive genitive should allow the elided noun to occur if they are indeed instances of ellipsis. I think the noun cannot appear, however:

 5a. *Ich kenne einen Fussballspieler der Fussballspieler. 
      'I know a soccer player of the soccer players.'

 5b.  Ich kenne einen der Fussballspieler. 
      'I know one of the soccer players.'

In sum, given such observations that ellipsis is not actually occurring in cases of so-called N-ellipsis, there is little motivation to view forms like einer in the example sentence as the indefinite article; it is, rather, an indefinite pronoun.

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.