Genealogical research is often complicated by variant spellings of family names. This is especially true in records previous to the 20th century since literacy rates were much lower than today. It's not at all unusual in the 19th century to find close to a dozen various spellings of a family name in records.
One method that has been used to minimize the problems that this creates for genealogists is Soundex coding. The basic priniciple of Soundex coding is that the letters are grouped into six groups based on patterns of interchangeability in spelling variants.
|Group 1||B, F, P, V|
|Group 2||C, G, J, K, Q, S, X, Z|
|Group 3||D, T|
|Group 5||M, N|
|Ignored||A, E, H, I, O, U, W, Y|
The Soundex code for a name consists of the first letter of the name followed by a hyphen and then the group numbers of the first three letters that are not ignored. For instance, STEWART becomes S-363, since T is in group 3, E, W, and A are ignored, R is in group 6, and T is in group 3. Notice that the common variants, such as STUART and STEWARD also are coded ase S-363.
If two letters from the same group are next to each other, the number of the second letter is not used. In SCHUHMACHER, since S and C are next to each other and are in the same group, the C is ignored, so SCHUHMACHER is coded as S-526 instead of S-252. Notice that the variants SHOEMAKER, SHUMACKER, etc., again all have the same Soundex code.
If there aren't enough significant digits to produce three numbers, zeroes are added on the end. For instance, BUCH becomes B-200. Note that the variants BUCK, BUX, BOOK, BOOKE, BOOCK, BOOCKE, etc., all have same Soundex code.
Since sorting surnames by Soundex coding instead of by actual spelling places most variant spellings together, it is frequently used to index and/or search genealogical records.