Is an expression like "How long does this letter take to arrive at London" unnatural?


(Excuse for my broken English) I am a Japanese student studying English.

The other day, my classmate was asked to translate the below Japanese sentence into English by the teacher:

  1. この手紙がロンドンに着くまで、どのくらい時間がかかりますか?

The teacher showed the answer like this:

  1. How long does it take a letter to reach London?

But my classmate had translated it into this:

  1. How long does this letter take to arrive at London?

Is this sentence #3 unnatural?

No one who was in the class can answer whether this sentence is natural or unnatural. If this sentence is unnatural, what is it that makes it unnatural?

Asano Ryo

Posted 2014-11-07T13:43:53.557

Reputation: 153



"A letter" means letters in general. It means, "If someone (not necessarily me) sends a letter to London, how long will it take to get there?"

"This letter" means a specific letter (like Japanese これ[kore], I think). "The letter that I am talking about (or pointing at or holding in my hand right now) - how long will this specific letter take to get to London?"

When we're talking about a specific letter, the verb "to do" needs to be in the future tense ('will'), because the letter hasn't yet arrived in London.

And as quant points out, you need to use 'in London' rather than 'at London'. Because London is a big place and the letter will be within it.

So it should be:

How long will this letter take to arrive in London?

if I want to know the travel time for this specific letter that I am holding in my hand right now, or

How long does it take a letter to reach London?

if I want to know the travel time for letters in general, not any particular letter, just a hypothetical letter that anyone might send.

How long does it take a letter to arrive in London?

is also fine: 'to arrive in' and 'to reach' mean the same thing in this context.


Posted 2014-11-07T13:43:53.557

Reputation: 370

The main point (when asking about a specific letter we'd use future tense) is more or less an "absolute". But although *arrive in London* is far more likely, Google Books estimates 168,000 things have *arrived at London* in print. It's not "wrong", just less common.

– FumbleFingers Reinstate Monica – 2014-11-07T14:10:11.620

@FumbleFingers: Interesting point. I think it depends on the size of the object and whether there is more than one possible destination for it within the destination town/city. For example, if Dundee has only one station and I'm on the train to Dundee, then I might hear the announcement "In 20 minutes this train will arrive at Dundee". But for London they would specify the station e.g. "... this train will arrive at London Paddington". London Paddington is one of a number of possible stations in London whereas 'Dundee' is the only station of that name, which makes it a destination in itself. – A E – 2014-11-07T14:20:00.297

"Passenger services at Dundee are provided by First ScotRail, CrossCountry and East Coast." vs "London is the centre of the National Rail network, with 70% of rail journeys starting or ending in London.". So actually it may be because Dundee Station is the name of a station whereas London is not.

– A E – 2014-11-07T14:23:07.893

Yeah - I quite agree that there are certain specific contexts where *at* becomes more likely. And by implication, there are other contexts where it's so unlikely it wouldn't be unreasonable to say it's "wrong". – FumbleFingers Reinstate Monica – 2014-11-07T14:50:51.327


  1. How long does it take a letter to reach London? -- [teacher's version]

  2. How long does this letter take to arrive at London? -- [classmate's version]

The teacher's version (#2) sounds good. That is, it is acceptable and is standard English.

Your classmate's version (#3) sounds unusual, and is probably unacceptable. (Though, some of the problem might be due to the subject "this letter".)

Let's look a bit at your teacher's version #2:

  1. How long does it take a letter to reach London? -- [teacher's version]

But let's insert the word "for" in there to see how it looks:

  • A. How long does it take [for a letter to reach London]?

This version #A basically means the same thing as version #2 (the teacher's version).

I'm going to spend a hunk of the analysis on that version #A, because what I want to show will show up cleaner this way, or so I hope.

Let's now look at version #A and some versions that could correspond to it:

  • B. It takes five days [for a letter to reach London]. -- [sentence is a declarative clause]

  • C. It takes how long [for a letter to reach London]? -- [interrogative clause, with interrogative phrase in situ]

  • D/A. How long does it take [for a letter to reach London]? -- [interrogative clause, with interrogative phrase fronted]

Notice that version #D (which is identical to #A) has subject-auxiliary inversion, which was done due to the fronting of a non-subject interrogative phrase ("how long").

In version #D, we can see that the pronoun "it" is the subject because it had undergone subject-auxiliary inversion and also because it is sandwiched between the auxiliary verb ("does") and the lexical verb ("take"). That the word "it" is sandwiched like that definitely shows that the word "it" is the grammatical subject of the main clause.

Let's look a bit deeper at version #B. We see that the subject is the dummy pronoun "it", and that the infinitival clause "for a letter to reach London" is sorta like a semantic-subject for the sentence--this is why this pronoun "it" is considered to be a dummy pronoun, because it has no semantic meaning in itself. In other words, version #B has the meaning of the below version #E, but #E is less likely to occur than version #B:

  • B. It takes five days [for a letter to reach London]. -- [extraposition]

  • E. [For a letter to reach London] takes five days. -- [infinitival clause as subject]

Version #B is called an extraposition construction because there's an extra position in the clause, one that is filled by that extra word "it", when compared to the more basic version #E.

The reason why extraposed version #B is more likely to be used than #E is that #E's subject is a subordinate infinitival clause, which means that the subject is a clause and that means that the subject will usually be relatively heavy in weight when compared to the other elements in a sentence, and there is a strong preference to have heavy stuff moved to the end of the sentence, thus, the extraposed version is often preferred (by the speaker/writer).

Now, having gone through all that, let's go and look at the teacher's version #2:

  1. How long does it take [a letter] to reach London? -- [teacher's version]

We can see that the word "it" is the grammatical subject of version #2. That's because the word "it" is sandwiched between the auxiliary verb ("does") and the lexical verb ("take").

So, there is a lot of similar stuff between version #2 and the above heavily analyzed version #A. But there is a big difference: in version #2, the so-called semantic-subject is the noun phrase "a letter" (while #A's semantic-subject is the infinitival clause "for a letter to reach London"). We can see this if we replace the dummy pronoun "it" of #2 with that noun phrase "a letter":

  • 2.s How long does [a letter] take to reach London?

Also, because version #2s doesn't lose any semantic meaning due to the deletion of the word "it", that indicates that the word "it" is a dummy pronoun (i.e. it has no semantic meaning).

Let's look at some versions that could correspond to version #2s:

  • 2.t [A letter] takes five days to reach London. -- [declarative clause]

  • 2.u [A letter] takes how long to reach London? -- [interrogative clause, with interrogative phrase in situ]

  • 2.v/s How long does [a letter] take to reach London? -- [interrogative clause, with interrogative phrase fronted]

Those above versions (#2s-v) seem to be okay. The difference between them and the previous versions (#A-D) is one that is more of pragmatics--in this case, of information packaging. That is, the subject in #A-D is a dummy pronoun "it", and so, it will be relatively quite easy to use their sentences in many different contexts -- while versions #2s-v have a semantically meaningful noun phrase as their subject and so there will be more restrictions on their sentences as to which contexts they will be acceptable in.

Now let's look at your classmate's version #3:

  1. How long does [this letter] take [to arrive at London]? -- [classmate's version]

Compared to your teacher's version #2, there are three main differences:

  • version #3 does not have a dummy pronoun "it" as its subject. Instead, it has a semantically meaningful noun phrase "this letter" as its subject, which means that the contexts that its sentence will be acceptable in are restricted.

  • version #3 uses a different noun phrase as its meaningful semantic-subject: #3's "this letter" (a definite noun phrase) vs #2's "a letter" (an indefinite noun phrase).

  • version #3 uses a different verb ("arrive") in its infinitival clause; and also the infinitival clause uses a preposition phrase ("at London") as a complement instead of a noun phrase ("London") as done in version #2.

Your classmate's version #3 is awkward and probably unacceptable. As to why this is so, it might help if we fiddle around with permutations of version #3, by switching out the three differences in different combinations to see which differences seem to be important.

For instance, if we switch in the verb "reach" (which means we'll also have to exchange out the preposition phrase for the noun phrase "London" to make it grammatical), we'll get:

  • 3.a How long does [this letter] take to reach London?

It sounds a little better than before, but it seems to be unacceptable, that is, awkward. The remaining awkwardness might be due to the noun phrase "this letter". (Notice that the awkwardness disappears once we switch in the noun phrase "a letter", but then that version has now become version #2s.)

Let's now try putting that permutation #3a into a construction that uses a dummy pronoun "it" as its subject:

  • 3.b How long does it take this letter to reach London?

It sounds a little bit more better, but it is probably still too awkward sounding to be acceptable.

Let's try a permutation where we put your classmate's original version #3 straight into a construction that uses a dummy pronoun "it" as its subject:

  • 3.c How long does it take [this letter] to arrive at London?

It too sounds a bit better than your classmate's original version #3.

Let's try permutations of your classmate's version #3 where we switch in the noun phrase "a letter", with and without using a dummy pronoun "it" as subject:

  • 3.d How long does [a letter] take to arrive at London?

  • 3.e How long does it take [a letter] to arrive at London?

Those changes seem to help a bit, but both of them still seem to be awkward and unacceptable.

So, there seems to perhaps be multiple reasons why your classmate's version #3 ends up being awkward and unacceptable.

  • One major reason is the difference in type of noun phrase (NP): the definite NP "this letter" used in your classmate's version #3 versus the indefinite NP "a letter" as used in the teacher's version #2.

  • Another major reason is the difference in the verb in the infinitival clause: the verb "arrive" in your classmate's version #3 versus the verb "reach" in the teacher's version #2.

Interestingly, notice how the acceptability of your classmate's version #3 seems to be improved if we insert the verb "will":

  • 3.f How long will this letter take to arrive at London?

  • 3.g How long will it take this letter to arrive at London?

Those versions sound better, but they are probably not yet (fully) acceptable.

Let's insert a "for" into version #3g to turn it into an extraposition construction (#3h), and see how it sounds:

  • 3.h How long will it take [for this letter to arrive at London]?

That version #3h sounds better to me, but it is probably still not yet (fully) acceptable.

As to which version would correspond better to the Japanese version #1, that's something that I can't help you with.


Posted 2014-11-07T13:43:53.557

Reputation: 5 118

+1 But ...... Really? *Semantic-Subject*???!!! I bet that hurt ;) – Araucaria - Not here any more. – 2014-11-09T12:51:13.400

1@Araucaria Yes, I had to stop and think a bit about it at the time--though, I thought my solution to be a good one as the term "semantic-subject" allowed me to side-step the entire thingie about syntactic functions versus semantic roles. I was also quite proud of the way I side-stepped the thingie about "it" and the NP "a letter" in the construction of "#2. How long does it take [a letter] to reach London?" – F.E. – 2014-11-09T23:03:50.397

Which of the couple or three it-thingies are you thinking of? I suppose the one where you don't mention that [a letter] and [to reach London] are two separate complements of take, and also what the function of take technically is? Yes, that was pretty crafty ... :) [not as crafty as your minorly outrageous explanation of extra position though] – Araucaria - Not here any more. – 2014-11-10T01:08:22.073

@Araucaria Oh, I just noticed a slip with me using the term "extraposition", and just fixed it. (Hopefully that was the only one, but I only skimmed the post this last time.) – F.E. – 2014-11-10T04:50:14.240

Aaargh, what happened with your comments?! I wasn't able to give them a good read, was being domesticated .... Oh well. Thanks for the helpful and insightful input. Re extraposition, didn't H&P 'steal' that term from TG? It's a bit strange, cos they have extraposed subjects, which makes 'extrapose' quite verby ... – Araucaria - Not here any more. – 2014-11-13T00:14:50.263

In that post there's no opportunity for whiz deletion is there, because left would be a complement of BE, not a modifier of dinosaur skin, right?

– Araucaria - Not here any more. – 2014-11-13T00:51:17.853

@Araucaria I'm kinda thinking now that the word "left" seems to be a past-participle verb form. E.g. "There were skins left (by the teacher) (for us) to study." – F.E. – 2014-11-13T01:41:18.993

@Araucaria So, that seems to support two internal complements, one as a "displaced subject NP" and the other as an infinitival extension. That is: "There were [skins left] [to study]." This seems similar to CGEL page 1394, [12.ii] "There were [some letters *written by her grandmother]* in the safe"--which uses a locative extension. – F.E. – 2014-11-13T01:47:48.223


@Araucaria A typical ELU thread full of grammar nonsense <== "a “noun phrase” is a phrase headed by a noun, while “NP” includes phrases of other sorts which can act in the same syntactic role as noun phrases." Lordy. I pity the students. (It doesn't help when the OP doesn't know the difference between direct and indirect objects.)

– F.E. – 2014-11-13T02:06:59.947

@Araucaria So, to answer your question about a possible whiz deletion. Using my previous example "skins left (by the teacher)", with what seems to be a bare passive "left", could be interpreted to have the semantic meaning of "skins [which were left] (by the teacher)" -- which is that, er, whiz-deletion. – F.E. – 2014-11-13T04:11:21.180

@Araucaria Your analysis concerning "There is no dinosaur skin to study left" doesn't seem solid to me (the example sounds awkward to me). I think "dinosaur skin left" is a nominal, and for it to be transformed into your version (with "left" postposed in the clause), that would involve moving "left" out of the NP and up to clause level; this is something that is very restricted even with full relative clauses in an ordinary sentence, much less now with a bare relative (which also is within an existential sentence). – F.E. – 2014-11-13T06:16:59.100

@Araucaria Oh, wait! That's an ELL thread! I can discuss grammar in an ELL thread! :) Maybe I should write an answer post there, but, wait, an answer has already been marked as accepted. Oh well. (snicker, snicker) Besides, I've already written two long answers this week--and one was very, very long. – F.E. – 2014-11-13T06:29:00.560

@Araucaria At the bottom of your post was [No dinosaur skin was left to study] <== Couldn't that have as its existential counterpart be "There was no dinosaur skin left to study"? (Which is similar to the OP's original example, but with "was" instead of "is".) – F.E. – 2014-11-13T07:12:32.037

@Araucaria In your last comment in that thread, there was this: In 'there were three plates left' it looks like it's postmodifying 'plates', but it isn't. <-- But I'm kinda thinking that "left" might be postmodifying "plates", because if it wasn't, then you'd have a "passive existential construction"! That is, "There were left three plates", which looks on the surface like a main clause passive, but supposedly, the NP "three plates" was postposed -- But remember example [14.ii] on page 1395, "There were killed [some 650 infantry from the 2nd Battalion]." (cont.) – F.E. – 2014-11-13T07:28:22.997

@Araucaria (cont.) CGEL says that the heavy NP can be postposed because "killed" is not a modifier of the NP! It calls "killed" an extension. And so, perhaps the word "left" is a past-participle extension in the OP's example (just like "killed is an extension in CGEL's example). Which seems to have the example on page 1395, [13.ii] "There were [several people] killed" being similar to the OP's example, except the OP's example has two extensions. Yes, I think this is getting closer! :) – F.E. – 2014-11-13T07:32:49.487

@Araucaria So, in conclusion (for now, anyway), it seems that the OP's example has a "displaced subject NP" and two extensions: "There is [no dinosaur skin] [left] [to study]." Which is sorta like the existential BE having 3 internal, er, "complements". Notice that it seems that there are two possible ways for the NP to be postposed: "There is [left] [no dinosaur skin] [to study]" and "There is [left] [to study] [no dinosaur skin]", which don't seem to be as good as the original. – F.E. – 2014-11-13T07:40:25.797

@Araucaria And now, I think that: There are two potential interpretations/parsings: 1.) two internal dependents (with "left" as a modifier in NP); 2.) three internal dependents (with "left" as its own extension). It might even be possible that both interpretations are allowable in the OP's example. – F.E. – 2014-11-13T09:35:21.137

That's good, because I came to that conclusion myself yesterday night. Aaaarrr I rewrote the whole darn post and then when I tried to save it our internet went on the blink! Yes, I reckon the most natural reading's 2 extensions. It definitely seems more like 13 ii than 12 ii to me. Although it also seems like it could be a predicative extension. Although it goes through with the two verb tests it seems like some of the items in 10 which also are quite restricted, I'm thinking available, vacant and in particular missing which is kinda like an opposite of left. – Araucaria - Not here any more. – 2014-11-13T09:36:32.613

I'm sure there's at least 2 readings! You still up?! I thoguht it was like 5am there or something like that! Thanks for all of that - very helpful indeed. and although I'd kind of got there, I was feeling a bit insecure in my evaluation! – Araucaria - Not here any more. – 2014-11-13T09:38:14.007

1@Araucaria ELL has better "saving" than ELU, where you might be able to actually recover your work (I was able to do that for an in-progress creation of an answer, which I thought I lost, but it was there when I tried to find it). -- It is 3:40 AM, and I'm going back to bed; I was lying in bed thinking things over, solidifying stuff and what not. bye. :) – F.E. – 2014-11-13T09:40:54.560

I'm not sure about left as NP modifier because it seems the counterpart's No dinosaur skin left is to study seems a bit wonky to me . What do you think about that? – Araucaria - Not here any more. – 2014-11-13T09:41:10.507

1@Araucaria I had thought about that; it might depend more on the existence of the distribution of "No dinosaur skin left" in other types of sentences. And that there might happen to be a restriction on that specific example, so that "is" can't be placed there. Which could give weight to the 3 dependent parse, but doesn't necessarily eliminate the allowability of the 2 dependent parse--or so my thinking was going like that earlier when I was supposed to be sleeping. Bye for real. :) – F.E. – 2014-11-13T09:48:49.740

@Araucaria After thinking some more about it, it seems that both interpretations can be shown to be supportable: that "left" can be an internal modifier in NP, or that "left" can be a clausal extension. They will provide two different semantic meanings, with two different non-existential glosses/paraphrases; the existential clause is ambiguous, and can be used in the two different contexts, where it will provide the two different meanings. – F.E. – 2014-11-13T19:14:14.167

1@Araucaria For a similar example: "There were (some) bones left (for us) to study". The two meanings: 1) Some bones were left on the table by the teacher for the purpose for us to study them; 2) Some bones happened to be left in the cave by time and floods and if we want to we can study them. – F.E. – 2014-11-13T19:17:37.110

Hmm will ponder, am out right now. Seems to me that left is the kind of complement that's only really licenced by BE and HAVE and a few other verbs - unless there is also a locative complement too. The following don't feel to happy: I saw some bones left; I liked those bones left. Bones left are a nuisance, so I'm not entirely sure aboutbeing a modifier in NP, because the acceptability seems to rely more on it being a/in a complement of certain verbs (leading me to believe that it is a complement not an adjunct, and of the verb not the NP), but am full of beer and not at home! :) – Araucaria - Not here any more. – 2014-11-13T21:06:13.980

Actually, saw some bones left looks ok, but the others are still squiffy, imo. – Araucaria - Not here any more. – 2014-11-13T21:09:24.333


Is this sentence unnatural? No one person who was in the class can't answerd. 
And if this sentence is unnatural, Where is it that makes this sentence unnatural?

Yes, it does sound unnatural. In English, we generally use at to describe specific objects that describe a location:

I arrived at the airport.

I arrived at home.

The letter arrived at the post office.

Locations, such as cities or countries, are generally referred to with in:

I arrived in New York.

I arrived in England.

The letter arrived in London.

That last one should answer your question :)


Posted 2014-11-07T13:43:53.557



How long does this letter take to arrive at London?

This sounds unnatural to me, because there is no way of knowing how long this letter will take to arrive in London. Perhaps it will never arrive. Maybe the postman will lose it.

How long does it take a letter to reach London?

This is much better. We can analyse how long letters take to reach London, and say, on average, it might take five days. We cannot, of course, say how long a particular letter will take.

How long does this letter take to arrive at London?

If you were going to use this construction (eg. at a Post Office) I might ask:

How long will this letter take to arrive in London?

They might reply:

Probably it will take four to five days.

Nick Gammon

Posted 2014-11-07T13:43:53.557

Reputation: 927


I think the problem here is much simpler than others have made it out to be.

Arriving happens instantaneously.

I arrived at 2 o'clock.

When will you arrive? (In an hour.)

Because of this, it's unnatural to ask how long something will take to arrive. Other verbs would suit better, like the other one in your example.

How long will this letter take to reach London?

How long will this letter take to get to London?

Your original sentence is understandable, and by no means would I stop you mid-sentence to correct you, but (because you asked), this use of arrive is a bit odd.

I would change it if I were editing a formal English document, for example.

As mentioned in other answers, at is also not the correct preposition to use. Usually, in is used with large bodies, like cities and countries. For example, I would arrive in Oslo, but at my hotel, since a hotel isn't miles and miles around.

Azor Ahai -- he him

Posted 2014-11-07T13:43:53.557

Reputation: 654