On Fri, 17 Apr 2009 08:55:19 +0200, Andreas Johansson <[log in to unmask]>

>On Fri, Apr 17, 2009 at 6:04 AM, Roger Mills <[log in to unmask]> wrote:
>> Certainly there is an element of transitivity in English and other
langs.' prepositions, in that they "take an object" just like a transitive
verb. But in my experience only English (maybe Germanic langs.?) can use
them intransitively i.e. as "particles" (and not all can be so used), and
IMO such prepositions may have changed class to become adverbs. Or maybe
they're adverbs to begin with, and can be used prepositionally??)
>I incline to the analysis that Germanic "prepositions" are better seen
>as particles that can work syntactically as adpositions, adverbs, etc,
>than as a separate word class whose members get zero-derived into
>adverbs, etc. 

I think it's the right analysis that prepositions can be used without an
object ("intransitively") in English, and that a whole lot of words
traditionally counted as both preposition and adverb, and even some that are
traditionally just adverbs, in fact are in fact simply preposition. 
"Adverb" traditionally is a poor catchall of a class; just jamming
everything in which can serve an adverbial function doesn't make for a very
homogeneous word class, or one about which you can say very much, so I'd
rather take things out.

This is the analysis that Huddleston and Pullum adopt in the _Cambridge
Grammar of the English Language_, and I made a post a month or two ago
maladroitly defending it which I can't find now.  

In fact, it's kinda curious to me that this analysis seems questionable to
so many people.  Is this another effect of the hapless word "preposition"? 
It's called a preposition, it has to be positioned pre something.  I mean,
I've not run across anyone who denies verbhood to intransitive verbs, or
even zero-valent verbs (like weather verbs in many langs); the case at hand
seems pretty similar (though, admittedly, verbs have their characteristic
morphology to help recognise the class).