# Origin of theta-joins

I've taken an introductory database course where like many good database courses, they taught relational algebra, including $\theta$-joins. However, I've recently read E. F. Codd's 1970 paper, A Relational Model of Data for Large Shared Databanks, and the concept of $\theta$-joins is nowhere to be found. He does use the symbol $\theta$ in the paper in section 2.2 though, but to refer to something much more general than joins.

In fact, Natural Joins are even defined slightly differently then the definition I'm familiar with. It requires that the original relations be completely recoverable from the joined relation.

So my question is when/in what paper/by who was the modern notion of the $\theta$-join introduced that is often taught by universities?

(A link to Wikipedia's definition of a $\theta$-join, similar to the one I learned in my course)

-