Subject question to subject attribute

Suppose that I know what a subject question is. Now, there's a more tricky case:

How many people came to the meeting?

The above one doesn't have an auxiliary verb did because it's a question to subject attribute how many (or a question to subject definition. frankly, I don't know how to name this grammar element properly)

Now, there's the other one:

How much time does it take to get from A to B?

How much time it takes to make a coffee?

So here is the question: which of the sentences above is grammatically correct? Why should I use does in the former, but omit it in the latter sentence?

And, finally, how should I write the following question:

How much time this deployment takes?

or

How much time does this deployment take?

Thanks!

Answer

SHORT VERSION:
What you want to say is

How much time does it take ... ?

with subject/auxiliary inversion and do support because the interrogative expression How much time refers to the missing object of the verb take. You say

How many people came ... ?

without inversion or do support, because the interrogative expression refers to the missing subject of the verb came.

LONGER VERSION:
_{Warning: what follows is not a full technical explanation. I've worked through a handful of those, and to tell the truth they make my head explode, and I think they would probably make yours explode too. So this is just a shorthand version or 'baby rule'; but I think it's all the ordinary learner needs to get a handle on the matter.}

Let's start by looking at the underlying declarative sentences which give rise to your contrasting sentences, with X representing the unknown value you're asking about:

1. How many people came to the meeting? ← X people came to the meeting.

and

2. How long does it take? ← It takes X {minutes/hours/days}

In 1), X is an attribute of the subject (this, I take it, is what you mean by "a subject question"); in 2), X is an attribute of the object.

Now let's look at the mechanics of inversion and do support in questions.

• 
  

  First the simpler process: polar (yes/no) questions, where you're not asking for specific value such as "five people" or "three hours" but for your hearer's assertion or negation of the proposition. In this case, the question is marked by subj/aux inversion: the subject changes places with the first auxiliary in the verb group. If there is no auxiliary, the appropriate form of do is inserted as a 'dummy' auxiliary (and the main verb is recast in the infinitive as its complement) to make the inversion possible: AUX present - no do support: No AUX present - do support required It has taken three hours ... It took three hours ... __| inversion ______| inversion + 'do' ↓ ↓ ↓ Has it taken three hours ...? Did it take three hours ...?

  
  
  


• 
  

  Now the more complicated process: variable questions like your examples, where you're asking for a specific unknown value. Here there is an additional step before inversion: interrogative replacement. The unknown value is replaced with a wh- expression, and that is 'fronted'—placed at the front of the clause to signal the question—leaving a 'gap', something missing in the body of the clause: It has taken [X time] It took [X time] ____________________| replace & front___________________| ↓ ↓ How long it has taken _ How long it took _

  
  

  Then inversion is carried out as before: How long it has taken _ How long it took _ __| inversion ______| inversion + 'do' ↓ ↓ ↓ How long has it taken _ ? How long did it take _ ?

  
  


• 
  

  This works pretty straightforwardly with variable questions where the unknown value involves an object or complement or adjunct, and even (with some extra complications) when it involves the verb. But when the unknown value involves the subject of the clause something odd happens. The first step, interrogative replacement, works as it did before [X many people] came to the meeting ____________________| replace & front ↓ ↓ How many people _ came to the meeting

  
  

  But on the next step, inversion, we all of a sudden come to a screaming halt—there's no subject left for the auxiliary (or dummy 'do') to invert with!

  
  

  So what do we do? . . . We do nothing. We leave the question as it stands, without inversion or do support:
  How many people came to the meeting?


  
  
  

  Why? . . . Well, the point of inversion and do support in other questions is to clarify the syntactic relationships: the subject is recognizable as the constituent which follows the auxiliary, so we know we're looking somewhere else for the 'gap' which the interrogative expression points to. But in this case it is the absence of the expected inversion which provides the necessary information: it tells us immediately that the 'gap' occupies the subject position.