I agree that the FRBR-LRM text could be clearer.  Maybe it comes down to this: when two nomens in the same system context are identical, when are they the same nomen and when are they different nomens?

Case 1.

Smith, J. [person A]
has alternate nomen - Smith, John

Smith, Johnny [person B]
has alternate nomen - Smith, John

Case 2.

Work A
has subject - Gardening [nomen for res C]

Work B
has subject - Gardening [nomen for res C]

I'd argue that in case 1, "Smith, John" and "Smith, John" are identical but different nomens because they link to different res, and the case 2 "Gardening" and "Gardening" are the same nomen because they link to the same res. But I've been wrong before. 

Stephen

On Mon, Mar 28, 2016 at 2:29 PM, Heidrun Wiesenmüller <[log in to unmask]> wrote:
Stephen,

I agree that a nomen cannot exist on its own, but must always be tied to an entity. But I don't think that two nomens are distinguished from each other simply by the fact that they are tied to two different entities. That is to say, I do understand *your* reasoning, but I cannot find any evidence in LRM that they see it this way.

The definition of nomen is "a designation by which an entity is known" (p. 20). And in the scope note says: "A nomen is whatever appellation is used to refer to any entity found in the bibliographic universe. (...) Depending on context of use, the same sequence of symbols can be assigned as a nomen of different entities in the real world even within the same language (polysemy and homonymy). (...) The association of nomens to entities is in general many-to-many". To me, this means that the same nomen (the same appellation) can be used - and in fact, often is used - for several entities.

Also, if being associated with different entities would automatically make the nomens unique, FRBR-LRM wouldn't have to state on p. 49: "In general, the appellation relationship would be many-to-many, however, in the context of a particular library system, the intention is that each nomen is used in an unambiguous sense by being associated with a single res."

But it is rather frustrating that we need to do, as it were, "exegesis" of FRBR-LRM. Couldn't the authors simply have stated clearly what they mean? The area of nomens is really rather fuzzy in LRM, especially as it includes unique identifiers as well as (potentially non-unique) names.

Heidrun



"An arbitrary sequence of signs is not a nomen until it is assigned to be an appellation for something in some context.  Only at this point is it significant in the bibliographic universe.  The attributes of the nomen entity actually characterize the relationship between the nomen and the thing (res) that is named." FRBR-LRM, Table 4.2, LRM-E9)

This stipulates that an unassigned string like "Smith, John" is not a nomen until it is assigned as a nomen for a particular John Smith.  That assignment serves to individuate the nomen.  The fact that many res entities might have the appellation "Smith, John" does not mean that there is a single nomen assigned to multiple res.  Rather, each nomen instance must be the appellation of a particular res (within a given system).  The fact that they all look the same is not relevant to the relationship each has with its res.  A name which is unassigned is not a nomen, and identical nomens assigned to multiple res does not imply the existence of an unassigned nomen entity, or a nomen independent of any particular res.

The difference between two nomens need not be categorical.  Simply being assigned as the appellations of two different res in the same category is sufficient to differentiate the nomens.

Stephen

On Mon, Mar 28, 2016 at 4:26 AM, Heidrun Wiesenmüller <[log in to unmask]> wrote:
Thomas,

I'm not worried about things like "780".

In the attributes list, there is "Category" for a Nomen (LRM-A25). I assume that we could record "DDC number" as a category. Although LRM  doesn't state this explicitly (at least I can't remember having seen this in the document), I also assume that there wouldn't be a conflict between two identical nomens if they belong to different categories. So there wouldn't be a problem in having 780 as a DDC number and having other cases of "780", as long as the categories are different.

Heidrun




On 27.03.2016 Thomas Berger wrote:
Am 27.03.2016 um 11:32 schrieb Heidrun Wiesenmüller:
Thomas said:


I don't see your problem here, nomens are much more than names, labels
or strings.
and

See above: There are many nomens with identical labels.
I'm sorry, but I don't see that at all. As far as I understand LRM, a
nomen doesn't *have* a label, a nomen *is* a label (name, title, string,
identifier...).  Look at the examples on p. 21f., p. 38 and p. 49.

On p. 20, LRM says: "Depending on context of use, the same sequence of
symbols can be assigned as a nomen of different entities in the real
world even within the same language (polysemy and homonymy). Conversely,
the same entity can be referred to by any number of nomens (synonymy).
The association of nomens to entities is in general many-to-many."
continuing

"An arbitrary sequence of signs is not a nomen until it is assigned to
be an appellation for something in some context. Only at this point is
it significant in the bibliographic universe.
***The attributes of the nomen entity actually characterize the
relationship between the nomen and the thing (res) that is named.***"
(emphasis mine).

one of the examples then is "780" (the DDC notation for "Music" as a
concept). Obviously there are many other things that could be named
"780", for instance the number 780. So there are many different
entities associated with "780".


This sounds quite alright to me. But on p. 49, LRM says: "In general,
the appellation relationship would be many-to-many, however, in the
context of a particular library system, the intention is that each nomen
is used in an unambiguous sense by being associated with a single res."
So it seems to me that it would not be acceptable for LRM to have "Peter
Miller" as a nomen of two different persons.
Indeed, and the prescription of the 1:m cardinality makes this
perfectly clear (if there was only one nomen "780" it would be
exclusively reserved to DDC usage in the bibliographic universe).

But it would be acceptable to have two different nomens colloquially
called "Peter Miller" (but would these be nomens for nomens?)

I think your quote of page 20 is crucial:

"Depending on context of use, the same sequence of symbols can be
assigned as a nomen of different entities in the real world even
within the same language (polysemy and homonymy)"

leaves doubt an the question how many nomens are involved when
e.g. there is one sequence of symbols and two different entities in
the real world.

also (the next paragraph):
"The identity of a nomen is given by the choice and order of the symbols
used within it"
could be interpreted that a nomen is uniquely identified (nomens of
nomens again?) by the choice and order of the symbols used within it,
i.e. there is only one nomen "780" in the universe. However in the
light of the succeeding sentence discussing variations in ordering and
variations in visual representation IMHO gives no occasion to back
up this interpretation.

I think the following sentence from p. 20 might be misleading:

"The association of a nomen to the entity it represents is a cultural or
linguistic convention, there is no inherent meaning embedded in the nomen."

could be interpreted as there were an abstract nomen "780" without
meaning and then different contexts go on and start associating
this with different res.

But I do not think this is admissible, it should be rather read as
follows: The nomen is necessarily and unvariably tied to exactly one res
(otherwise we could not consider it a nomen) but the meaning lies
in this tying, not in the symbols used.

I could concede that FRBR-LRM is not clear about nomens, and both
interpretations exist:

(1) There is only one nomen "780" in the universe
    => utterly unusable as a model

(2) There are (at least) as many nomen "780" in the universe as
    there are different entities named by "780"
    => might be of some value since these "individual" 780's
       can then be equipped with attributes describing the
       context of usage (a string "780" could not).

viele Gruesse
Thomas Berger







You have received this message because you are a subscriber to the  RDA-L discussion list. 
You can change the email associated with your subscription, the method that you receive RDA-L posts, and other settings by logging into your subscription at http://lists.ala.org/info/rda-l. 
To unsubscribe from this list, send an email to [log in to unmask] with UNSUBSCRIBE RDA-L in the subject line. Leave the body of the email blank. Once the email is received you will be automatically unsubscribed. Please note you will be unsubscribed for the address from where the email originated.





-- 
---------------------
Prof. Heidrun Wiesenmueller M.A.
Stuttgart Media University
Nobelstrasse 10, 70569 Stuttgart, Germany
www.hdm-stuttgart.de/bi



--
Stephen Hearn, Metadata Strategist
Data Management & Access, University Libraries
University of Minnesota
160 Wilson Library
309 19th Avenue South
Minneapolis, MN 55455
ORCID:  0000-0002-3590-1242


-- 
---------------------
Prof. Heidrun Wiesenmueller M.A.
Stuttgart Media University
Nobelstrasse 10, 70569 Stuttgart, Germany
www.hdm-stuttgart.de/bi



--
Stephen Hearn, Metadata Strategist
Data Management & Access, University Libraries
University of Minnesota
160 Wilson Library
309 19th Avenue South
Minneapolis, MN 55455
Ph: 612-625-2328
Fx: 612-625-3428
ORCID:  0000-0002-3590-1242