The STARTING WITH predicate tests whether or not the first characters of the first value expression match those specified in the second value expression. The STARTING WITH predicate has the following form: (B)0[m[1mstarting-with-predicate = [m [1m [m [1mqq> value-expr qwqqqqqqwq> [1;4mSTARTING[m[1m [1;4mWITH[m[1m qq> value-expr qq> [m [1m m [1;4mNOT[m[1m qj [m [1m [m Because the STARTING WITH predicate is case sensitive, it searches for uppercase characters and does not include lowercase characters for the DEC Multinational Character Set; the reverse is also true. For example, STARTING WITH 'Ç' retrieves a set of records different from those retrieved by STARTING WITH 'ç'. The STARTING WITH predicate is sensitive to diacritical markings used in any Multinational Character Set. Therefore, a matches a, but neither matches á, à, ä, Á, À, Â and so on. In Spanish, ch and ll are treated as if they were unique single characters. For example, if a domain is defined with the collating sequence SPANISH, then STARTING WITH 'c' does not retrieve the word char, but retrieves the word cat.