Why does "fragen" take two accusatives?

Question

Ich muss dich etwas fragen.
— Duolingo

(If the above is wrong, so is the premise of this question, so please correct it if need be!)

Does this sentence not have two direct objects? Why does the verb not take dir, i.e. the dative?

I'm comparing this with English ask or French demander : ask something (COD) of someone (COI).

This is not a usage question but an analytic one: Is there some argument or thematic scheme with which this is consistent across German(ic languages), and if so, what's the history behind it?

Answer 1

Most sentence plans follow a few general rules. If there are multiple actants involved, typically the actant with the most active role is in the nominative case (subject), the actant with the least active role is in the accusative case (~ accusative object) and a possible remaining actant is in the dative case (~ dative object). So typical plans would be

• K_sub (= the actant who "acts"), e.g.: Ich schlief.
• K_sub (= the actant who "acts") K_akk (= the actant who "receives"/the matter/object concerned), e.g.: Ich schlug sie. Ich kaufte eine Cola.
• K_sub (= the actant who "acts") K_dat (= the actant who "receives") K_akk (= the matter/object concerned), e.g.: Ich schenkte meinen Eltern eine Vase. Ich schrieb ihr einen Brief.

Needless to say, this doesn't hold true for all verbs. For instance, a very small subset of verbs follow the pattern

• K_sub K_akk (K_akk),

an example being

• jemand (= the actant who does the asking) fragt jemanden (= the actant who is being asked) (etwas [= the matter that is being asked about])

In some respects, "fragen" is even an outlier within this group of outliers. Also notice that the structure is rather firmly established in the case of "fragen", unlike in some of the other cases, which is why this valence pattern is commonly understood to be somewhat unstable. See, inter alia, Gallman in Duden - Die Grammatik, 8th ed., 2009, recital 1485). So what's going on with "fragen"?

Zifonun et al., Grammatik der deutschen Sprache, vol. 2, 1997, p. 1085, note that this "exceptional valence pattern" likely can only be explained historically (emphasis mine):

Bis ins Mhd. regiert fragen wie viele andere Verben des Bittens und Instruierens einen Akkusativ der Person und einen Genitiv der Sache. Während ein Großteil dieser Verben einen Rektionswechsel zu Dativ der Person und Akkusativ der Sache (jemandem etwas beweisen) erfuhr oder - bei Erhalt des personalen Akkusativs - im Nhd. der Gegenstand mit einer PP ausgedrückt wird (jemanden um etwas bitten, jemanden vor etwas warnen, jemanden für etwas loben, auch jemanden nach etwas fragen), scheinen bei fragen mit doppeltem Akkusativ sozusagen diese beiden Muster zusammengefallen zu sein.

Answer 2

There is no such thing as direct and indirect object in German. That's a concept from French which got shipped across The Channel. German instead has accusative, dative, genitive, and prepositional objects, and object clauses.

There's a small group of verbs which can take two accusative objects:

• fragen
• nennen, taufen, rufen, schimpfen, schelten
• lehren, unterrichten
• abfragen, abhören
• angehen
• kosten

Answer 3

We derive grammar rules from languages. Languages do not arise from rules. If languages don't "obey" rules, the rules are defective, not the languages. In addition, in the case of European languages, we try to impose rules derived from Latin onto languages that did not descend from Latin, which makes it even worse. We say:

• He is, Isn't he?
• You are, Aren't you?
• I am, Amn't I? (whoops)

"Amn't I?" isn't hard to pronounce, so why do we say "Aren't I?" Why is this not taught in secondary school as an ungrammatical irregularity? It is the only case in which we change the form of the verb to make a question. It's because English isn't Latin. German is also not Latin. Cramming them into Latin boxes doesn't always get the expected results.

We say:

• ich frage dich
• ich frage etwas

So it makes perfect non-Latin sense to say:

• Ich frage dich etwas