Suggestion: For lists of names, sort them alphabetically.

Rationale: It's simple, everybody understands it and the indexing program can do it automatically. If it is important for names with the same initials to sort together then simply omit the names:

So:

and these automatically sort together.

Suggestion: For lists of names, omit full stops between initials.

Rationale: They take up space and add nothing.

Recommended: Spaces in items are treated as a letter which precedes all others. This results in word-by-word? order.

Symbols are treated as preceding letters.

Digits are treated as following symbols and preceding letters.

Rationale: Users of indexes are literate and therefore familiar with the order of the letters in their alphabet. They can therefore quickly find a specific item in a long list.

Recommended: Spaces in items are treated as a letter which preceeds all others. This results in word-by-word? order.

Rationale: Word-by-word order is better at grouping together items of similar meanings and so is well suited to subject indexes.

Rationale: Letter-by-letter order groups items purely by their spelling and it can be quicker for the user to find a single item. This order is suitable when the items are not interconnected, such as in dictionaries.

Symbols are treated as preceeding letters.

Rationale: Users are more likely to look at the beginning of an index than the end and so will see that the index does include some entries for symbols.

Digits are treated as following symbols and preceeding letters. But where digits form numbers, then they should be placed in numerical order, so 10Base-T preceeds 100Base-T.

Rationale: Users are more likely to look at the beginning of an index than the end and so will see that the index does include some entries for symbols. Numerical order is more intuitive for numbers.

Rationale: This may group together similar items in a useful fashion. A user must find consistency of rules in an index to be confident that when an item is not found where it logically should be, then there is no information for that item on that topic.

Hyphens are treated as spaces.

Rationale: Hyphens are used to join together words in portmanteau fashion and so word-by-word sorting will sort under the first word, e.g Korea-Singapore Free Trade Agreement

Standard Variant CMS

Rationale: Hyphens are used to separate prefixes from words which would otherwise be confusing or difficult to pronounce and so the prefix is not a word in its own right.

• Smith, J.A.
• Smith, John Alfred
• Smith, J.O.
• Smith, James Owen

• Smith, J.A.
• Smith, James Owen
• Smith, J.O.
• Smith, John Alfred

Alphabetical Order

Commonly Agreed Standard:

Items are grouped together according to the way they are spelt and put in the order of the alphabet of the language of the index user. All items beginning with the same letter, whether uppercase or lowercase, are grouped together. Where multiple items begining with the same letter, items are ordered according to the next letter, and so on. Rationale: Users of indexes are literate and therefore familiar with the order of the letters in their alphabet. They can therefore quickly find a specific item in a long list.

Recommended: Spaces in items are treated as a letter which preceeds all others. This results in word-by-word? order. Rationale: Word-by-word order is better at grouping together items of similar meanings and so is well suited to subject indexes.

Alternative: Spaces in items are treated as not being present. This results in letter-by-letter? order. Rationale: Letter-by-letter order groups items purely by their spelling and it can be quicker for the user to find a single item. This order is suitable when the items are not interconnected, such as in dictionaries.

Symbols are treated as preceeding letters. Rationale: Users are more likely to look at the beginning of an index than the end and so will see that the index does include some entries for symbols.

Digits are treated as following symbols and preceeding letters. But where digits form numbers, then they should be placed in numerical order, so 10Base-T preceeds 100Base-T. Rationale: Users are more likely to look at the beginning of an index than the end and so will see that the index does include some entries for symbols. Numerical order is more intuitive for numbers.

Additional items (double entries/cross-references) may be created for numbers which are sorted as though they were spelt out. If this is done for any entry, it must be done for all numerical entries in the index. Rationale: This may group together similar items in a useful fashion. A user must find consistency of rules in an index to be confident that when an item is not found where it logically should be, then there is no information for that item on that topic.

Standard Variant ISO999: Hyphens are treated as spaces. Rationale: Hyphens are used to join together words in portmanteau fashion and so word-by-word sorting will sort under the first word, e.g Korea-Singapore Free Trade Agreement

Standard Variant CMS Hyphens are treated as not being present. Rationale: Hyphens are used to separate prefixes from words which would otherwise be confusing or difficult to pronounce and so the prefix is not a word in its own right.

Suggestion: For lists of names, sort by Surname/Family Name and then by Initials of Forename, regardless of whether they are spelt out.

Rationale: This makes it easier to find a person as there will be a single position in the index for them, regardless of whether their name is spelt out. Where there may be multiple entries for a single person, perhaps from the merging of multiple references, they will be brought together. So:

Smith, John Alfred Smith, J.O. Smith, James Owen

rather than:

Smith, James Owen Smith, J.O. Smith, John Alfred