Quoting Jonathan Rochkind <[log in to unmask]>:

> But if we don't have any accessible means of mapping from LCC to  
> LCSH...   then it's not an option. So do nothing, or do something  
> else?
> Hmm, what if you displayed the LCC headings in the shelf browse,  
> _and_ included them in the index as searchable, to meet the "if you  
> display it, the user should be able to search it" rule.  But of  
> course you don't stop indexing LCSH either -- LCC is a bad 'entry  
> vocabulary', we don't expect anyone to use it 'naturally' -- but if  
> they see it on the shelf browse and then want to use it to perform  
> searches, at least they can.

I don't have access to LCC so I don't know what their "labels" look  
like on the class numbers. I have only vague recollections based on  
the examples in the MARC classification format. So it isn't clear to  
me if they will be good ones to show users. I am pretty sure that the  
records contain a portion of hierarchy that provides context, e.g.  
(from the online classification MARC standard):

$aPQ4315.25$hItalian literature$hIndividual authors.$hIndividual  
authors and works to 1400.$hDante Alighieri,  
commedia.$hInferno.$jParticular cantos

Since both LCC and LCSH are "syndetic" -- that is, the actual  
instances get built from a set of rules and lists -- no version of  
either standard will be complete (although I hear that LC is trying to  
include more pre-composed instances in the version of LCSH,  
probably for this reason). This means that what you have in the  
record, in many cases, will not match what is in the standard. That's  
why I though FAST would be interesting because it could include those  
standard lists. But I am also told that has a lot of those  
standard lists in it... Which means that, as usual, you've got a bit  
of a mess -- some of your headings/class numbers will match, some will  
need an add-on to match, it's not clear if you can identify the  
added-on portion, especially in LCC, because the class numbers are not  
hierarchical the way they are in Dewey.

Another note/oddity is that LCSH has what they call "pattern" headings  
-- so there are few people included in LCSH (people are in the LC  
Names file) but you can find Shakespeare because he's the pattern for  

and there are others who are patterns for historical figures, etc. (I  
think George Washington is one.) So again you've got something that  
isn't complete for matching except in a few instances. (All other  
authors MAY have the same subheadings as Shakespeare, but they won't  
be in the subject file; there will just be a name in the LC Names file.)

Essentially, the digital versions of LCSH and LCC derive from the  
instruction tomes that were used by subject catalogers. They aren't a  
complete list of either the call numbers nor the subject headings, but  
they explain how to create instances of LCC and LCSH. For this reason,  
although it is great to have them in a machine-readable form, it  
doesn't get you what you would often expect from a controlled  
vocabulary list, which is a complete list of the vocabulary that you  
can use for matching.

Another note (then I'll shut up): In LCSH cross references are not  
made to all of the headings that could use them, it's kind of a  
cascading thing. This is made up, but it works like this:

   use: United States

But no:

USA. Post office
   use: United States Post office

The entry vocabulary assumes that you have started alphabetically and  
have "seen" the reference. If you want to provide that capability in a  
system you have to create your own xrefs using all of the possible  

Now I'm really depressed, and I'll stop.


> It would certainly be better if we had one vocabulary we could use  
> for all things that serves all purposes, but it's just not the  
> current environment. Even if you did have an accessible means of  
> mapping from LCC to LCSH (I don't think anyone other than maybe OCLC  
> does; and I am not sure if even OCLC shares it with members in a  
> flexible machine readable way, unless maybe the Terminologies  
> Service does), even if you did have that... I wonder, if LCC works  
> better hieararchically for a shelf browse label, if you might still  
> want to show both LCC and LCSH as labels in a shelf browse.
>> Quoting Jonathan Rochkind <[log in to unmask]>:
>>> Ah, but we're not talking about "entry vocabulary", we're talking  
>>> about labelling shelf ranges.
>> At my job at UC we had a rule: if you display it, the user should  
>> be able to search it and get those same results. If you display one  
>> set of strings as a shelf label, and a different set of strings are  
>> required for retrieval, that's going to be confusing. Ideally users  
>> should be able to search within the classification scheme, or to  
>> navigate around, but we don't have that ability. The point of my  
>> blog post is that we have separate systems and it can't be clear to  
>> users how they interact. (I'm not even sure they do interact  
>> cleanly.) The only way users can make sense of things is by  
>> extrapolating from what we display to them. I worry that seeing  
>> inside LCC, while being given only LCSH to search on, isn't going  
>> to be clear. While LCSH loses a lot of the structure of LCC, at  
>> least users are seeing what they would need to search on in the  
>> catalog to get those same results.
>> kc
>>> But I agree that the headings for LCC will be less user friendly  
>>> than LCSH.  If there was a way to get LCC-to-LCSH mappings in an  
>>> easily usable way without paying tens of thousands of dollars,  
>>> that would be clever. (I'm not sure there's a way to get them even  
>>> if you DO pay millions of dollars).
>>> So I was suggesting using the LCC headings themselves as a more  
>>> feasible alternate plan, is all. I agree it would be insufficient  
>>> if we needed an "entry vocabulary".  But just for labelling shelf  
>>> ranges on display, I think it's probably not worse than nothing.
>>> Of course, that's up to the implementer, what's better than nothing.
>>>> Quoting Jonathan Rochkind <[log in to unmask]>:
>>>>> For #2, you can provide a useful topical/subject type heading  
>>>>> via much simpler and more feasible solutions than mapping to  
>>>>> LCSH.  For #2, you don't need a map to LCSH, you need the LCC  
>>>>> schedules with descriptions of what each range of LCC call  
>>>>> numbers is for, in machine-readable form.
>>>> I would give the opposite advice. LCC will have fewer  
>>>> pre-composed headings than LCSH at, and the  
>>>> terminology associated with the numbers in digital LCC will be  
>>>> less user-centric than the LCSH subject headings. cf my most  
>>>> recent blog post:
>>>> There isn't any entry vocabulary for users other than LCSH --  
>>>> which isn't really entry vocabulary to LCC and is definitely NOT  
>>>> entry vocabulary to DDC.
>>>> kc
>>>>>> Thanks everybody!
>>>>>> this is useful for a couple of purposes
>>>>>> 1) sometimes we have records that have call numbers, but no  
>>>>>> subject headings.
>>>>>> this would be useful to provide those.
>>>>>> 2) i'm thinking of providing a 'subject heading' label to our  
>>>>>> shelf browser --
>>>>>> so users see, in addition to the callnumber -- what the call  
>>>>>> number means.
>>>>>> thanks again!
>>>>>> rick
>>>>>> On Tue, Oct 18, 2011 at 2:33 PM, Jonathan  
>>>>>> Rochkind<[log in to unmask]>  wrote:
>>>>>>> Anyone know if the OCLC Terminology Service provides such a  
>>>>>>> mapping? The
>>>>>>> Terminology Service may be free if you are already an OCLC cataloging
>>>>>>> member.
>>>>>>> At one point I think I saw an absolutely free open access  
>>>>>>> machine readable
>>>>>>> mapping somewhere, that was made at some point in the past and  
>>>>>>> no longer
>>>>>>> updated... but I cant' remember where I saw that even.
>>>>>>>> LC's Classification Web provides a mapping from LC classifications to
>>>>>>>> LC subject headings.  There is a manual web interface, used mainly by
>>>>>>>> catalogers, which requires a subscription:
>>>>>>>> I don't know if it has any kind of API.
>>>>>>>> Keith
>>>>>>>> On Tue, Oct 18, 2011 at 2:11 PM, Enrico Silterra<[log in to unmask]>
>>>>>>>> wrote:
>>>>>>>>> is there any way to go from a LC call number,
>>>>>>>>> like DF853  to
>>>>>>>>> via some sort of api? opensearch?
>>>>>>>>> thanks,
>>>>>>>>> rick
>>>>>>>>> -- 
>>>>>>>>> Enrico Silterra Software Engineer
>>>>>>>>> 501 Olin Library Cornell University Ithaca NY 14853
>>>>>>>>> Voice: 607-255-6851 Fax:     607-255-6110 E-mail: [log in to unmask]
>>>>>>>>> "Out of the crooked timber of humanity no straight thing was  
>>>>>>>>> ever made"
>>>>>>>>> The information transmitted, including attachments, is intended only
>>>>>>>>> for the person or entity to which it is addressed and may contain
>>>>>>>>> confidential and/or privileged material. Any review, retransmission,
>>>>>>>>> dissemination or other use of, or taking of any action in reliance
>>>>>>>>> upon, this information by persons or entities other than the intended
>>>>>>>>> recipient is prohibited. If you received this in error,  
>>>>>>>>> please contact
>>>>>>>>> the sender and destroy any copies of this document.

Karen Coyle
[log in to unmask]
ph: 1-510-540-7596
m: 1-510-435-8234
skype: kcoylenet