Relational (inner) joins are commonly used in databases. However, there are different definitions and implementations of joins. One way to understand a join is to think of it as “looking something up” or adding redundant data to existing data. For example, when normalizing tables, we create a separate table for related data and use a join to retrieve the original table. Joins can also be seen as nested loops over rows or columns, compatible alternate realities, flatMap operations, solutions to the N+1 problem, paths through a graph, minimal models, typechecking operations, operations in the Set monad, or the biggest acceptable relation. By understanding these different perspectives, we can gain a deeper understanding of joins in databases.
https://justinjaffray.com/joins-13-ways/?a=b