tagSets G and M have been revised, based on these principles. See TagSet-G - Revised TagSets G and M.
For example, the Museum community can use the Title element interoperably with loose semantics; within that community they all know what they mean by a Title. Similarly, the Genealogy community uses Title as well, but with different implied semantics. The problem arises when the Museum of Genealogy want to provide records. Then they will need a schema that defines the semantics of Title.
Thus a tagSet-G element may be attributed stronger semantics when it occurs within the context of a specific schema. For example, the element Author has generic semantics such that if it occurs outside the context of a schema it is interpreted as "Author or Creator". A specific schema (for example, pertaining to museum objects) may confine its meaning to "Creator".
For the remainder of this document, an element occurrence is referred to as "context specific" or "generic" depending respectively on whether it occurs within or not within the context of a specific schema.
For example, suppose a client executes a search across databases creating a result set that represents records from potentially several domains. When the client retrieves a record from the result set, unless the retrieval record includes a schema identifier, the client might not know what schema governs the interpretation of the elements of the record. Suppose in this case the first several elements are from tagSet-G, say: Author, Title, and subject. Following these initial tagSet-G elements, assume there is a structured element with subelements, the first of which is a schemaIdentifier (tagSet-M element 1). Suppose that the client does not support the identified schema, and suppose further that there are several tagSet-G elements following the schema identifier. The client is able to discern that the Author, Title, and Subject pertain to the record, but the client is not able to process the record further, in particular, the client may not infer information from the second set of tagSet-G elements. In this case, the client may be able to inform the user that there is a potential record of interest, and that in order to interpret the record, support for the specific schema is necessary.
As another example, consider a database where an individual record corresponds to a physical object (for example, a work of art) and the record includes one or more digital renditions of the object. An abstract record structure may specify that tagSet-G elements may occur at the beginning of the record, outside the context of a specific schema, followed by a schema identifier, followed by a repeating, structured element, with a repetition for each rendition. There may be tagSet-G elements within each such "rendition" element, and these would pertain to the specific rendition. (The 'Title' or 'Author' for the first rendition may be different from those of the 'second' rendition. For example, the first rendition may be "black and white still image, 35mm" where the listed 'Author' is the photographer; while the second rendition may be a sketch of the physical object, where the listed 'Author' is the artist of the sketch. The Title and Author in both cases would be different from those of the object itself.)
A client receiving a record that includes these GILS elements (assuming the GILS schema to be in effect) should not infer any information from the presence of these tagSet-G elements, unless the client supports the GILS schema.
Library
of Congress