We are not talking about laundry lists though. In the context of
research outputs in a repository, I think we have a right to assume
that normal editorial standards would insist that the content of a
paper is comprehensible, and that if the sum total of all the words in
the paper don't represent what the paper was about then there would
have been a violation of the usual rules of writing.
Certainly one can make arguments about terminologies and language
drift and implied terms, but I don't think that they are core.
--
Les
On 25 Jun 2008, at 15:35, John Smith wrote:
> Les,
>
>> On 25 Jun 2008, at 13:53, Peter Cliff wrote:
>>
>>> I'm not convinced the content of a document is enough to
>> accurately
>>> (and sustainably) describe what it is about.
>> I think that you are taking a remarkably bold epistemological
>> position
>> there.
>
> What is so bold about it? All he is saying is that he is not
> convinced that the contents of a document (with no knowledge of
> context) are enough to reliably tell us what it is about. This seems
> intuitively true. If I were given a laundry list from the records of
> a 400 year old great house I doubt I would recognise half the items
> listed without access to other documents.
>
> I feel it is you who is bold in challenging him.
>
> We have been here before and when I asked if your algorithms could
> tell me what the Waste Land (Eliot) was about or even if it was a
> poem you replied you could tell it was a poem because the writer was
> a poet and it was published in a poetry journal. In other words you
> needed to look at the context.
>
> Simple document analysis (without context) has its place as a tool
> of the information specialist but it is only one tool amongst many.
> On its own it is has limited value.
>
> Regards,
>
> John.
>
|