Grammatical titbit

[Richard Hudson]

That's interesting Tom. I must admit I hadn't noticed it either. I think 
the examples you're trying with zero subject relatives are a red 
herring, because your main point applies when the relative is inside a 
subject NP:
(1) The man I think stole my bike is over there.
(2) **The man stole my bike is over there.
But the explanation is surely quite simple: at the point just after 
"man" where the speaker chooses between zero and who/that, the only 
thing that's relevant is that the pronoun would NOT be subject of the 
*next* verb, "think", which already has its own subject ("I"). In other 
words, the pronoun would be just extractee, not subject or object, in 
relation to "think". It's only further into the sentence that its 
relation to "stole" becomes relevant. So the rule for using zero 
pronouns (or whatever you want to call them) is that they're allowed 
unless the understood pronoun would be subject of the first following verb.

Does that make sense?

Dick


Richard Hudson www.phon.ucl.ac.uk/home/dick/home.htm

On 16/09/2011 15:19, Tom Bartlett wrote:
> Here's an interesting titbit I haven't seen picked up on before:
> With mental processes as pseudomodals in relative clauses THAT can be omitted even when it is acting as Subject:
> That's the man I think stole my bike.
> *That's the man stole my bike.
> My first thought was that this was simply because "gardenpathing" caused by the juxtaposition of Subject and Finite had been disrupted, but the same isn't true with modal adverbs:
> *That's the man possibly stole my bike.
>
> This doesn't seem to be down to the influence of the congruent use of the projecting verb either:
> *That's John; I think stole my bike.
> Does this jar with anyone's idiolect?  Or, conversely, is anyone happy with:
> *That's the man stole my bike.
> Is this possible in Northern English English?  I am tempted by other sentences such as:
> ?You're the one told me to do it!
> Any ideas?
> All the best,
> Tom.
>
>