Related: Niemand oder Niemanden?

According to Wiktionary and one or two other sources I looked at, niemand and jemand are either uninflected or take an -em ending in the dative case. I was checking the DWDS usage database to see which is more common (too close to call), but it turned up examples with -en as well. Here are some from Die Zeit:

Sorgen hatte sie sicher nicht, wer hat die schon, aber dass man existiert und schön ist und nett und alles, ist doch erst wahr, wenn es bei jemanden in den Augen steht.
Es macht natürlich nervös, mit jemanden zu wetteifern, den man nicht einordnen kann.
Unterschätzt wird Nahles in der SPD allerdings von niemanden.
Diese helfe niemanden.

Are these examples correct, incorrect, colloquial, or am I getting the case wrong? If correct, is the -en always a possible ending for the dative of jemand/niemand?

  • I'm puzzled concerning the uninflected statement in Wiktionary. If used as a real pronoun, I would always use the inflected form and restrict the uninflected variant to phrases as jemand Unbekannten, jemand Anderen.
    – guidot
    Feb 16, 2022 at 8:35
  • 2
    @guidot I think, it is actually not the uninflected form, but the Dativ/Akkusativ inflection in the process of loosing their endings. Most famous example: Keine Macht für Niemand. But in everyday language I hear constructions like Ich habe niemand gesehen all the time. Feb 16, 2022 at 8:54
  • 1
    @jonathan.scholbach They're not losing their endings. The endings on jemand and niemand first appeared in Neuhochdeutsch, as this answer explains
    – Numeri
    Feb 16, 2022 at 11:08
  • @guidot: By uninflected I mean having no -en or -em ending. Also from Die Zeit: Das würde er nie tun bei jemand, den er für einen Versager hält. Wiktionary says using the inflected forms "is probably somewhat more common (at least in writing)." The fact that the uninflected form is used in front of adjectival nouns was mentioned in Wiktionary and some of the grammars I looked at, for example LEOs deutsche Grammatik
    – RDBury
    Feb 16, 2022 at 11:10
  • 1
    @Numeri Thank you for the hint - very interesting! You are right, they are not even losing their endings, the -em and en ending for Akkusativ and Dative are maybe not even fully established. The comment for guidot remains, that this does not mean it was an uninflected form. Feb 17, 2022 at 8:15

3 Answers 3


My inclination is to dismiss von niemanden, mit jemanden and the like as errors. I cannot recall ever having heard such forms; the dative either has no ending or the ending ‑m. One would assume that the source of the error lies in the fact that the letters m and n are adjacent on keyboards and that it is a mistake that spell checkers will not point out.

Having said that, the ending ‑n in the dative occurs with such surprising frequency that the Duden grammar explicitly mentions it under paragraph 424:

Im Dativ tritt außerdem die Endung ‑en auf, zwar seltener als ‑em, aber doch in recht hoher Frequenz, wie eine Korpusuntersuchung des IDS zeigt.

The study mentioned is reported in https://grammis.ids-mannheim.de/fragen/4332. With bei, mit, von, zu, forms with ‑n instead of expected -m occur with a frequency of at least 1 in 15, going up to 1 in 10.

Adding to the mystery is the fact that the ending ‑n in the dative was possible historically. DWB jemand has two examples (under 3b):

von iemanden anders erfahren haben

acht tage habe ich dazu einen ausschlag über den ganzen körper gehabt, dasz ich mich kaum vor jemanden sehen lassen konnte

The possibility of having ‑n in the dative is mentioned by Hermann Paul, Deutsche Grammatik, vol. 3, § 138:

Die Flexion war im Mhd. die eines regelmäßigen Substantivums, also G. iemannes, D. iemanne, A. ieman. Diese Flexionsweise setzt sich bis auf den heutigen Tag fort, nur daß sich für den Dat. ausschließlich die Form mit Abwerfung des e festgesetzt hat. Aber daneben ist seit dem 17. Jahrh. adjektivische Flexion aufgekommen, A. niemanden, D. niemandem und niemanden, Formen, die jetzt ebenso üblich sind wie die älteren.

The ending ‑n could be interpreted as a weak adjective ending or a weak nominal ending; the fact that it occurs at all is mysterious. (In other contexts, ‑n can substitute for expected ‑m, but only when preceded by another word ending in ‑m; see the Duden grammar under paragraph 1527.) Given the fact that most modern grammars do not mention this form at all, I would treat occurrences of ‑n in newspapers as simple mistakes.

  • So the upshot, as I see it, is that it's an error but a common one that's been documented. And while it seem to have some historical precedent, it is, at best non-standard in Modern German. I take your point that spell checkers would not catch this; being a horrible speller myself I'm very familiar with English misspellings that spell checkers miss (capital/capitol, canon/cannon, etc.) I'd be surprised if a grammar checker didn't catch it, but I suppose they aren't foolproof either.
    – RDBury
    Feb 16, 2022 at 21:01
  • I updated the relevant usage notes in Wiktionary. There is still a bit of a mystery about why this is so common, but I'll leave that as a problem for the grammar police. I was mainly concerned that there might be something I was missing, a plural form for example. If it's just an error which, for whatever reason, happens to be common, then that's all I need to know.
    – RDBury
    Feb 16, 2022 at 21:37

All of the examples are wrong. Dative case is required for

bei jemandem

mit jemandem

von jemandem


jemandem helfen

There is no dative form of jemand ending with -en.

  • 6
    Jemanden would be a correctly formed accusative, however, as in Ich sehe jemanden.
    – guidot
    Feb 16, 2022 at 8:31
  • 2
    I wouldn't have asked if this occurred only a few times. But mit jemanden appears 29 times in Die Zeit, and 20 times in Berliner Zeitung. This is only about 10% the rate that mit jemandem occurs, but I would have thought that editors of major newspapers would have caught this kind of thing nearly all the time.
    – RDBury
    Feb 16, 2022 at 11:29
  • 2
    It's a very common error in German and it even happens to quality newspapers. Feb 16, 2022 at 13:34
  • @Thomas Wana: I guess it's an error, but frequent enough that it's starting to border on an accepted variation. If enough people do it wrong then it becomes allowed but not preferred, if more people do it then it becomes preferred, and if still more do it becomes mandatory. Thus the language evolves.
    – RDBury
    Feb 16, 2022 at 19:54
  • @RDBury It's not something, that becomes allowed because language evolves. It's clearly an error and I'm really surprised (if not shocked) that it happens so often in a newspaper like Die Zeit. Jemanden is accusative. Maybe it's because of jemand that can be used for dative and accusative and somehow the -en gets added to avoid using jemand? I don't know.
    – Olafant
    Feb 16, 2022 at 22:01

There is no situation in which jemanden can be used after a dative preposition. If this occurs in text, it's a typo or a case of faulty OCR or transliteration of a source. The only valid forms of jemand in the dative case are jemand itself or jemandem.

Interestingly, as an aside, jemand predates jemandem, with the latter not occuring until New High German, which started around 1650 A.D. This answer (not mine) explains that in more detail.

With that aside out of the way, I have to admit that your question made me curious. I've never seen someone make a mistake like that in writing – or maybe I've just never noticed!

Using the DWDS corpora, I searched for all occurrences of the form <dative-preposition> jemand|jemandem|jemanden. Unfortunately, I couldn't get the boolean OR operator working well, so I had to search for each combination separately. To recreate my data, make sure to search using the syntax @mit @jemanden, which ensures exact matches for each word.

The following table shows the relative frequency of each version of jemand after all dative prepositions which make sense in conjunction with jemand (I did not include seit or ab, for instance). It also includes the total number of such constructions found.

While it's still certain that these occurrences are errors, I do find it interesting that they occur so commonly in the DWDS data.

Preposition jemand jemandem jemanden Total
aus 23.53 % 52.94 % 23.53 % 34
außer 85.71 % 0 % 14.29 % 7
bei 16.53 % 76.15 % 7.32 % 369
gegenüber 35.71 % 57.14 % 7.14 % 28
mit 17.27 % 73.71 % 9.02 % 1963
nach 15.38 % 77.83 % 6.79 % 221
von 28.72 % 61.36 % 9.92 % 2047
zu 24.32 % 64.21 % 11.48 % 366


The raw data I collected (I excluded certain corpora to avoid low quality sources or what might have been repeated data, i.e., the metacorpora):

prep pronoun DWDS-Kernkorpus (1900–1999) DWDS-Kernkorpus 21 (2000–2010) DTA-Kernkorpus (1598–1913) Berliner Zeitung (1994–2005) Der Tagesspiegel (ab 1996) Die ZEIT (1946–2018) Blogs
aus jemand 1 0 0 2 1 3 1
aus jemandem 1 0 0 4 2 7 4
aus jemanden 2 1 0 1 0 2 2
außer jemand 0 0 0 2 0 0 4
außer jemandem 0 0 0 0 0 0 0
außer jemanden 1 0 0 0 0 0 0
bei jemand 13 1 3 4 6 13 21
bei jemandem 17 8 1 34 37 126 58
bei jemanden 0 0 1 1 1 5 19
gegenüber jemand 1 0 0 1 0 7 1
gegenüber jemandem 2 0 0 1 1 10 2
gegenüber jemanden 0 0 0 0 0 1 1
mit jemand 65 3 80 18 17 80 76
mit jemandem 125 43 3 195 162 688 231
mit jemanden 3 1 27 20 9 29 88
nach jemand 8 2 4 3 4 11 2
nach jemandem 15 4 0 25 18 84 26
nach jemanden 1 0 3 4 1 2 4
von jemand 74 8 182 32 27 141 124
von jemandem 81 25 9 159 110 564 308
von jemanden 7 1 83 4 9 13 86
zu jemand 17 3 19 4 2 32 12
zu jemandem 20 10 3 31 30 104 37
zu jemanden 3 0 5 3 2 9 20
  • 1
    I'd throw any queries with less that 5 result; there's often some twist of phrasing that allows two words to meet in some unexpected way. Your results do prove that the inflected form is more common than the uninflected form, something was unsure of from a visual inspection. They also support what I was saying that the en ending is occurring more often then I'd expect for a simple typographical error.
    – RDBury
    Feb 16, 2022 at 20:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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