On 7/27/2011 3:49 PM, john g marr wrote:
> On Wed, 27 Jul 2011, Jonathan Rochkind wrote:
>
>> I thought 045 was used for "aboutness" time period? Like a
>> historical book about the 19th century, or perhaps a novel that takes
>> place in the 19th century -- time periods covered by the content itself.
>
> "Aboutness" is a sufficiently subjective concept that it can be
> interpreted broadly. For example, anything written in Plato's time can
> be considered to be representative of the time and its thought, and
> thus be "about" that time.
Hmm, we're failing to communicate here. Sure, you _could_ conceptualize
things that way. It's just that it wouldn't solve the problem that you
were proposing it to solve.
The original problem identified was a user who wanted: "Show me all the
things originally written in year X-Y, (regardless of when the
particular reprint was published, I want the stuff originally written
conceived in that range.)" Right?
So, sure you could conceive of something written in Plato's time to be
"about" that time. But the problem posed was not a user wanting
everything "about" that time. It was a user wanting things originally
written in that time.
If you record Historical works about Plato's time but which were written
in the 20th century a "Plato's time" 045, and you ALSO record things
originally written in Plato's time a "Plato's time" 045....
...then we can now solve a different user problem that nobody had
suggested was a problem in a first place, the problem "Show me
everything about Plato's time (and I consider things written in that
time to be about it to)".
But we still can't solve the problem that was posed, that you were
proposing could be solved, but which can't be. The system still can't
assemble a set of things originally written/conceived in Plato's time.
Therefore the data is just as useless as it was before for solving that
problem.
Do you understand this?
This exchange is probably an example of the difficulty of catalogers and
programmers to communicate, I guess. It is probably frustrating for us
both, we seem to be talking about different things while trying to talk
about the same thing.
Sure, you can add more data in a way that is a nice internally
consistent "can be considered" thing that might make sense to you. But
if doesn't allow the system to answer the question you thought you were
providing for, the question that you know or believe that users want
answered, it's useless. The point isn't just putting the data in the
record somewhere that can be justified 'can be considered' logically
coherent; it has to be putting the data in the record in a way that the
system can answer the questions you want the system to be able to
answer. If metadata designers can't think in those terms, the metadata
will not be functional.
Received on Wed Jul 27 2011 - 17:35:39 EDT