Each question relates to the German sentence just above it. I hope these apparently five questions are merely asking the same question five times.

  1. die Bohnen fünf Minuten kochen lassen

Is die Bohnen the object of kochen or lassen?

  1. Ich lasse mich nicht belügen.

Is mich the object of belügen (in which case it is an indirect reciprocal) or of lasse (in which case a direct reciprocal)?

  1. Auch einer der Männer, die neben dem Wagen gingen, hatte sich abhängen lassen und ging jetzt auf einer Höhe mit mir.

Is sich the object of abhängen or lassen?

  1. sich einen Bart/die Haare wachsen lassen

Is sich the dative complement (or indirect object) of wachsen (someone allows: the beard growing to himself) or the dative complement of lassen (someone allows to himself: the beard growing)?

  1. Man fragte ihn von Rechtswegen, was er [Candide] lieber wolle: sechsunddreißigmal durch das ganze Regiment Spießruten laufen oder sich auf einmal zwölf bleierne Kugeln durchs Hirn jagen lassen.

Is sich the dative complement of jagen (Candide allows: someone’s driving bullets into his brain) or the dative complement of lassen (Candide allows to himself: bullets being driven into the brain)?


To answer the question, you don't have to read this background or any postscript.

In English, we may have letting someone come or letting someone cook the beans but not letting cook the beans. We might however hear, Well, let it cook for five minutes, in which case cook suddenly becomes intransitive with the same meaning as be cooked (in which cooked is transitive).

But not every English verb will accommodate this switching. For example, we would not say Let Tom call in the sense of Tom rufen lassen. (I am assuming that Tom rufen lassen is having someone call Tom. Letting Tom call will always mean letting Tom call someone else.)

So I am wondering which of the following is the case:

(a) In German, any typically transitive verb can become implicitly passive and thus intransitive in a lassen context. In die Bohnen fünf Minuten kochen lassen, for example, a German speaker will see letting the beans cook (be cooked) for five minutes (which has essentially the same structure as ihn kommen lassen). Likewise, in ihn rufen lassen, a German speaker will see letting him be called.

(b) In German, we allow lassen to take an infinitive with only an object (patient) but no subject (agent). In die Bohnen fünf Minuten kochen lassen, therefore, a German speaker sees letting an unspecified agent cook the beans for five minutes (or allowing the cooking of the beans). In ihn rufen lassen, he or she would see letting an unspecified agent call him (or allowing the calling of him).

Note: By speaking of “what a German speaker sees” I am not trying to confine meaning to psychology. I will be equally happy to receive native speaker intuition or any other relevant support for (a) or (b). (For instance there might be evidence by analogy to another language in which the direct and the indirect reciprocals look different.) Of course, if neither is right, please let me know what is.


I am realizing that 4 might actually have these many potential syntactic analyses because a question arises not only in respect of sich but also of Bart.

(a1) lassen[sich, Bart] wachsen[-, -] (allow to himself a beard's growing)

wachsen is passive and intransitive. sich is the dative complement of lassen. Bart is the accusative object of lassen but the agent of wachsen.

(a2) lassen[-, Bart] wachsen[sich, -] (allow a beard's growing to himself)

wachsen is passive and intransitive. sich is the dative complement of wachsen. Bart is the accusative object of lassen but the agent of wachsen.

(b1) lassen[-, -] wachsen[sich, Bart] (allow growing a beard to himself)

wachsen is active and transitive. sich is the dative complement of wachsen. Bart is the accusative object of wachsen. (wachsen does not have an express agent.)

(b2) lassen[sich, -] wachsen[-, Bart] (allow to hisemlf growing a beard)

wachsen is active and transitive. sich is, however, a dative complement of lassen. Bart is the accusative object of wachsen. (Again, wachsen does not have an express agent.)

(b3) lassen[-, sich] wachsen[-, Bart] (allow himself growing a beard)

wachsen is active and transitive. sich is the accusative object of lassen and the agent of wachsen. Bart is the accusative object of wachsen. (I guess in this case we don't know whose beard we are talking about. Someone might have permitted himself the silly hobby of tending a goat's beard. Therefore, (b3) cannot be what a German speaker "sees" in a typical instance of sich einen Bart wachsen lassen.)


Why did I reverse the beans and cook to create an example of something not used in English, i.e. letting cook the beans?

I see in the comments section that at least three people have had this question. If you are having this question, it means indeed you have not grasp the very thing under consideration. So, well you might ask.

Very quickly, letting cook the beans is meant to be a schematic representation of a syntactic structure not used with English let, but common for German lassen, namely:

ALET + let + [null] + infinitive + PINF     :     e.g. you let [null] do it

where ALET is the agent (subject) of let, and PINF is the patient (object) of the infinitive.

One catch is that, because of the ordinary inversion of verb and object in a dependent context, German would have this surface structure:

ALET + lassen (lässt) + [null] + PINF + infinitive

An example of German in this structure is:

ihn rufen lassen

when this means having him summoned.

That it could have some other meaning not relevant to the topic (e.g. having him summon another) need not detain anyone.

  • @addy2012 Do you not see the difference between (a) and (b)? If (a), we have a single syntactic form but the infinitive (e.g. of cook or kochen) switches between active and passive in meaning. If (b), the verb is always in the active, but we get two syntactic structures. As you say, it is a question that can arise in English as well (only in English the answer is clearly (a)).
    – Catomic
    Commented Jan 14, 2016 at 15:22
  • ihn rufen lassen can mean both "allow him to call" as well as "Have someone call him"
    – adjan
    Commented Jan 14, 2016 at 15:31
  • "Lass ihn (doch) seine Oma rufen" can either mean "He should call his grandma" but also "Allow him to call his grandma". It's really context. Usually you can hear the difference. It has a very different tone to mean "allow" than "have him"
    – adjan
    Commented Jan 14, 2016 at 15:33
  • so "Lass ihn rufen" can mean: -"have someone call him" -"allow him to shout" -"make him shout"
    – adjan
    Commented Jan 14, 2016 at 15:37
  • "lassen" has so many different meanings: deutschegrammatik20.de/spezielle-verben/verb-lassen
    – adjan
    Commented Jan 14, 2016 at 15:41

3 Answers 3


In some Indo-European languages, there's a construction called Accusative and Infinitive (abbreviation AcI, Accusativus cum infinitivo). In this construction, the accusative object of the main verb also serves as subject of a dependent clause, and in some languages (e.g. Latin), if this object is a pronoun, the reflexive form is obligatory.

So, to answer your questions:

I. Does the accusative object in the example sentences belong to the main verb lassen, or to the dependent verb?

Let's see if we can change the sentences to include accusative objects for both verbs:

  1. Sie lässt den Mann die Bohnen fünf Minuten kochen.

  2. Sie lässt den Mann sie nicht belügen.

  3. Sie lässt den Mann sich vom Wagen abhängen.

  4. Sie lässt den Mann sich die Haare wachsen.

  5. Sie lässt den Mann sich zwölf Kugeln durchs Hirn jagen.

So you can see that in all sentences, the object of lassen is missing in the originals, and the objects that are present all belong to the dependent verbs.

II. Does lassen make the dependent verb passive?

In German, lassen certainly implies a passiv-ish mood. Among the English translations of lassen there's also leave, so they idea is that some other action happens without the main subject actually doing anything.

As you have seen, it's true that the direct object of lassen can be implicit. However, I wouldn't go as far as to say that this changes the dependent verb into the passive form, or a transitive verb into an intransitive one (on a grammatical level as opposed to expressing a mood).

Reflexive pronouns in general also can make an action "passiv-ish", even though grammatically it is not. So there's (semantically) a continuum between active and passive. This may be connected to the fact that e.g. in Latin, the reflexive pronouns are always used for the AcI.

I also really couldn't distinguish between (a) "leave the beans alone so they cook" and (b) "letting an unspecified agent cook the beans". These are really two sides of the same thing, which is expressed by lassen.

III. How to analyze "sich den Bart wachsen lassen"?

Just like in example sentence (2), sich can in principle be both be the accusative object of lassen and the dative object of wachsen. So it's ambigous, or both, and I don't think trying to distinguish transitivity or intransitivity helps. Additionally, sometimes a reflexive construction is halfway between transitive and intransitive anyway, so it's complicated enough even without lassen.

  • Thanks. But I don't understand why, in section I, 2 would come out any different than the others. The object present in the original, i.e. mich (which has become the second sie in your re-write), would be that of the dependent verb belügen. The (accusative) object of lassen is the one you supplied, i.e. den Mann. Therefore, 2 is no different than the others, and the (accusative) object of lassen and that of belügen are not identical. No?
    – Catomic
    Commented Jan 15, 2016 at 14:32
  • The "idea is that some other action happens without the main subject actually doing anything." - I don't think we are talking about the same thing. Suppose we let Caesar conquer the world. Yes, some action (the conquest) happens without our doing anything further; so that may be "passivity" in your sense. But that doesn't taint the verb conquer with any implicit passivity in the grammatical sense (as happens when we let the beans cook for five minutes).
    – Catomic
    Commented Jan 15, 2016 at 14:41
  • That you cannot distinguishing between (a) and (b). This is very informative. My question (obviously) tries to make sense of the behavior of infinitives in the lassen context with no more than ordinary grammatical concepts pertaining to verbs in any event (such as active vs. passive voice, subject and object). But native German speaking minds (I presume you're one) may have recourse to grammatical categories peculiar to the lassen context; hence the resistance to a reductionist account.
    – Catomic
    Commented Jan 15, 2016 at 14:51
  • @Catomic: I was trying to say that in principle you cannot distinguish if "mich" is the object of "lassen" or the object of "belügen". So it's not obvious that "mich" has become the second "sie". But you are right, OTOH "belügen" is always transitive, so some object is needed, and OTOH the one doing the lying cannot be "sie" herself, so in this case, it also must be the object of belügen. I'll edit the answer. There are examples like "ich lasse mich gehen" where the dependent verb is intransitive, so "mich" must be the object of lassen (resp. the subject of the dependent clause).
    – dirkt
    Commented Jan 16, 2016 at 10:38
  • @Catomic: I'm indeed not sure if we are talking about the same thing. We let Caesar conquer the world could either be translated as "Wir lassen es zu, dass Cäsar die Welt erorbert" (we allow through inactivity that Caesar conquers it) or as "Wir lassen Cäsar die Welt erobern" (that can has both the above meaning and "we told Cäsar he should conquer the world, and now we sit back and watch while he does it"). Yes, neither does taint the verb with any implict grammatical passivity. But"wir lassen die Bohnen kochen" doesn't taint "kochen" was passivity, either.
    – dirkt
    Commented Jan 16, 2016 at 10:45


Based on the responses so far it seems to me that the answer is "neither (a) nor (b)" insofar as interpretation is conceived of as a matter of "what happens in the mind of a native speaker."


In English, we have basically these two forms which let + infinitive may take, where A stands for agent and P for patient.

ALET + let + PLET/AINF + infinitive + PINF     :     e.g. you let me do it

ALET + let + PLET/AINF + infinitive + [null]     :     e.g. you let me come [null]

What we don't have in English is the following.

ALET + let + [null] + infinitive + PINF     :     e.g. you let [null] do it

But German allows what is equivalent to this third form, for example:

Herr Green empfing sie sehr eilig, als sei vieles einzuholen, nahm Herrn Pollunders Arm und schob Karl und Klara vor sich in das Speisezimmer, das besonders infolge der Blumen auf dem Tische, die sich aus Streifen frischen Laubes halb aufrichteten, sehr festlich aussah und doppelt die Anwesenheit des störenden Herrn Green bedauern ließ. (From Der Verschollene by Kafka.)

--where Anwesenheit is the patent of bedauern, but we have nothing to serve as the patient of ließ as well as the agent of bedauern.

Another example is:

ihn rufen lassen

but only if you take the meaning of this expression that helps to set up the question, i.e. having him summoned, instead of some other meaning that has nothing to do with the price of tea in China.

Looking at this third form, I believe a beginning student of German coming over from English has two strategies available to him (in a reverse order to the presentation in the main question).

(b) Simply accept that German has a third form (not found in English); that German speakers use lassen to mean what in English would be letting the calling of him and letting the regretting of someone's presence without expressly specifying who is to do the calling or regretting.

(a) Or alternatively do a mental trick to change the voice of the infinitive from active to passive so you see letting him be called and letting the presence be regretted.

Part of my question was to ask whether German speakers adopted either of these strategies. But I seem to be finding out that the "third form" is so natural to a German speaking mind that no strategy of any kind is necessary--even to the extent that the question does not even register.

  • As a student of German, (b) was covered on the most part by equating it to the construction "have/get sth. done [by so.]" (have the car fixed, get one's hair done, etc.), though admittedly, German is far more extreme in its Genauigkeit when it comes to specifying an extraneous agent: "einen Drachen steigen lassen", comes to mind. As for (a), this seems no different to me than replacing lassen with allow or as a friendly imperative, "let us pray", "let's go", etc. One form for which I did not find a similar structure in English is "etw. lässt sich [gut] + Infinitiv".
    – user19407
    Commented Jan 18, 2016 at 14:32
  • @user19407 I cannot understand your comments unless I assume you have (a) and (b) backwards. So assuming, yes, (a) and (b) are meant to be "obvious" strategies; so they should be familiar to you. As for etw. lässt sich [gut] + Infinitiv it has an exact analogue in English if sich is the patient of lassen and agent of the infinitive; alternatively, if sich is the patient of the infinitive, the sentence is simply an instance of the "third form" I am talking about. The presence of the reflexive does not make it a new structure.
    – Catomic
    Commented Jan 19, 2016 at 0:50
  • Yes, that looks similar to what I thought. And no, native Germans need neither of these strategies. It is often the case when two languages employ different constructs, to the non-native speaker it looks like there must be a "strategy" to wrap his head around this usage. For example, English has another AcI construction "I want him to do s.th.", but you need to replace that with a subclause in German "Ich möchte, dass er etwas tut". So when I learned English, I mentally replaced constructions like these until I got used to it, and now I can think directly in that way.
    – dirkt
    Commented Jan 19, 2016 at 11:22
  • BTW, the Latin AcI is much more similar to the German AcI (after all, Latin doesn't have fixed SVO order like English), and when I learned Latin in school, I never had any problems understanding these constructions directly, without the need for any strategies.
    – dirkt
    Commented Jan 19, 2016 at 11:24

The answer to all your questions is very simple. German has a complex predicate. You first combine the verbs into a predicate, and then you add the complements. The complements count as complements to the complex predicate, not to any single verb. This, at least, is the structure of the German clause (in contrast to the structure of English). At the meaning level, you can still try to figure out to which of the verbs the complement relates more closely.

Therefore, the German case is also different from the classical AcI construction. Latin AcI may involve a verb that embeds a clause with an accusative subject and an infinitival verb. The German structure involves only one clausal level.

For details: https://de.wikipedia.org/wiki/Koh%C3%A4rente_Konstruktion

Your Answer

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

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