Well, I should note that nouns ending with the letter S are considered plural in English. But here we deal with a Latin collation, and the Romans had their Latin plural nouns ending with 'ae', 'i', 'ia', 'es' and other stuff :hehe:
If we substitute 'peppers' by 'pepperl', that 'pepperl' still will be in the end of the list... So I don't think that plural is the explanation.
Latin1_General_CS_AS is Windows collation so it is follows the same word sort rules as a normal Windows string compare.
Check the property of this collation by executing the below sql
,COLLATIONPROPERTY(name, 'CodePage') AS CodePage
,COLLATIONPROPERTY(name, 'LCID') AS LCID
WHERE name = N'Latin1_General_CS_AS';
This query shows that the collation is case-sensitive, accent-sensitive, and the code page 1252 and LCID 1033, if you have locale as 1033, then default Windows string comparision rules will provide the same behavior as the database engine like any other in plain english.
The interpretation of the word "latin" does not indicated the real latin language here. 🙂 and the code page 1252 is like most used common code page in general everywhere. So I also added the soundex code to show how the value changes when you add "s" at the end where the consideration of the sort will change.
The first and the hardest SQL statement I have wrote- "select * from customers" - and I was happy and felt smart.