[Corpora-List] About Part of Speech in English and Chinese

[Taras Zagibalov]

The language is not static and it is not possible to assign a POS to
any sequence of characters constituting a word (whatever it means).
Nominalisation, polysemy, homonymy and other "semantic fluctuations"
will always make it difficult to attribute a word to a POS 'once and
forever'. Meaning is mostly contextual, only abstract part of it can
be stored in a dictionary (and POS is the most abstract layer of a
word's semantics), but such an abstract meaning does not exist 'on its
own' but only in contexts.

On Mon, Nov 2, 2009 at 2:37 PM, Mike Maxwell <maxwell at umiacs.umd.edu> wrote:
> Mike Scott wrote:
>>
>> (1) church tower
>> (2) tall tower
>> it is clear that (2) is adjectival, but in the case of (1) some linguistic
>> theories will call church a noun (because that word-form arguably is mainly
>> used for nouns) while others would call it an adjective because it is here
>> premodifying a noun. The former theories seem to act as if word-forms had a
>> primary POS, rather as people have their gender determined before birth,
>> while latter theories allow for the possibility that words may swing both
>> ways, so to speak, depending on the company they keep.
>
> Another difference between the two sets of theories is that the first theory
> says that phrase structure either allows for alternative "fillers", i.e. a
> rule like
>   NP --> Det {Adj|N}* N
> whereas the second theory says that in any given position, only one type of
> category can appear.  (I'm assuming a phrase structure theory here--I don't
> know enough about dependency grammar to say how this would carry over.)
>
> Actually, there is another analysis of (1) and (2) that is probably better:
> a rule like
>   NP --> Det Adj* N+
> (that's a flat version; one might have intermediate levels of structure).
>  This analysis would account for the following distinction:
>   a tall church tower
>  *a church tall tower
>
> Of course another distinction between 'church' and 'tall' in the above
> examples is that 'tall', but not 'church', allows pre-modification by words
> like 'very', 'really', 'extremely', 'slightly'.  I don't see how this result
> would follow if 'church' were an adjective in this environment.
>
>> The second aspect concerns the information supplied in the context or
>> inferable from it. In the case of (3) ... chief distribution ...
>> English simply does not tell us without more context whether we are
>> talking of the way chiefs (e.g. tribal chiefs) are distributed through a
>> population or territory, or whether we are talking of the main patterns of
>> distribution of something. Either way, chief premodifies distribution. In
>> POS tagging for such a case, the context may or may not disambiguate so POS
>> tagging will necessarily, for those linguists who think word-forms have a
>> predetermined POS, be varied.
>
> True.  A drawback from an engineering point of view, but not necessarily
> from a linguist's point of view.
> --
>   Mike Maxwell (the other Mike)
>   What good is a universe without somebody around to look at it?
>   --Robert Dicke, Princeton physicist
>
> _______________________________________________
> Corpora mailing list
> Corpora at uib.no
> http://mailman.uib.no/listinfo/corpora
>

_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora