Normalization

Normalization refers to the system processing whereby diacritics, most punctuation and special characters, and case differences are stripped from access fields for the purpose of providing strings for indexing and searching, and for comparing headings and references.

In LC's normalization process, subfield delimiters and the first comma in subfield a (for all types of headings and references) are retained. The ampersand (&) is also retained.

LC, RLIN, and OCLC have agreed to normalize headings and references in the same way, but there are some discrepancies. There may also be discrepancies between LC normalization and the same process in a given local ILS.
In all such cases, the LC normalization rules are followed in determining conflict in NACO. (For example, a hyphen normalizes to a space in all types of headings.)

The following forms are thus in conflict:

Large Scale Computer Analysis and Modeling Conference
Large-Scale Computer Analysis and Modeling Conference

Ku, Chun
Ku, Ch'ün

United States Information Agency
United States.‡bInformation Agency

Chung, Hui
Chung-hui

Johnson & Johnson
Johnson and Johnson