I'm struggling to come up with a definition of a "block" in a $(0,1)$-matrix $M$ such that we can decompose $M$ into blocks, but the blocks themselves don't further decompose. This is what I've got so far:
Given any $r \times s$ $(0,1)$-matrix $M$, we define a block of $M$ to be a submatrix $H$ in which: (a) every row and every column of $H$ contains a $1$, (b) in $M$, there are no $1$'s in the rows of $H$ outside of $H$, (c) in $M$, there are no $1$'s in the columns of $H$ outside of $H$, and (d) no proper submatrix of $H$ satisfies (a)-(c).
I want to define blocks in such a way that there are no proper sub-blocks in blocks. But I feel item (d) is difficult to parse. I can't just say "there's no proper sub-blocks" because this is a circular definition.
Q: Could the community suggest a better way of phrasing this definition?
My conundrum reminds me of this comic:

I would suggest breaking this up into two definitions... maybe call one of them a "block", and another a "simple block". That is:
How's that? Alternatively, call the first one a "pre-block", and the second a "block" (cf. pre-sheaf vs. sheaf).