# Poizat's definition of $p$-equivalent $k$-tuples

Maybe this is too trivial a question to be posted anywhere, but anyway. I am reading Poizat's "A Course in Model Theory". In page 4 he defines the notion of two $k$-tuples, each in the universe of some relation, being $p$-equivalent.

He gives two conditions:

1. $a_i = a_j \leftrightarrow b_i = b_j$.
2. The function $s$ defined by $sa_1=b_1,...,sa_k=b_k$ is a $p$-isomorphism from $R$ to $R'$.

I guess the first of them is just a remark, because it seems redundant given the second. If the function $s$ is a $p$-isomorphism then it is, in particular, a bijection between the finite subsets $\{a_1,...,a_k\}$ and $\{b_1,...,b_k\}$, and hence it is impossible to falsify condition 1. Am I missing anything?

-
Don't worry, we've seen more trivial questions :-) –  joriki Oct 19 '11 at 10:00

Yes, I was missing something pretty obvious. Condition 1 is what enables him to talk about the function $s$ in condition 2. Otherwise he should have talked about the relation $s$.