SPL has very different concept of JOIN from SQL.

Here is a brief introduction: https://c.scudata.com/article/1686883667617#toc_h2_19

For non-professional programmer(user of esProc desktop), we recommend the book: http://c.raqsoft.com/article/1638237297732

In A4, you should use field name like “countrylanguage.CountryCode”,“countrylanguage.Language”,“country.Name”,….

View A3, and you will see the field names of the table, there is no field named A1,A2.
A1,A2 is the intermediate variable name, after executing A3, they don’t exist in A3’s value.
The A1,A2 in A4 will be explained as the cell A1,A3, and A1.CountryCode means the A(1).CountryCode(first row). so the result of A3 would consist of many same rows, and then after grouping, A4 would have only 1 row.

BTW, xjoin is not the most appropriate function here, it runs slower.
You can use join:
A3= A1.join(CountryCode, A2:Code, Name, Populcation, Continent)
And write A4’s code according to the result in A3.

A4 is trying to perform grouping, but there is no aggregation expression, maybe you have not finished the code.

The datasouce is from MySQL. I think the calculation can also be done in one SQL statement:
SELECT … FROM country JOIN contrylanguage ON countrylanguage.CountryCode=country.Code