I work with someone on the core developer team and he pronounces it "see-ran," which is consistent with the explanation given by @cderv.
As for the pronunciation of the "pipe" operator (%>%), my guess is that the name comes from Unix/Linux shell scripting, where the pipe operator is the | (vertical bar). It has been a while, but when I worked exclusively in Unix terminals with a team of researchers, we would say, "pipe it to." For example, the Unix expression "ls -lt | more" would be "ls minus lt pipe it to more." So now I say "output pipe it to input" for "output %>% input" in R.