How to pipe text output to speech dispatcher?

I am not getting the command pipe to work for say and spd-say. Specifically, I want to pipe the output of fortune to spd-say, which seems more advanced than say. I tried the usual pipe construction fortune | spd-say Other commands do work, like fortune | wc Apparently, the speech dispatcher commands treat the piped input different from the ordinary arguments. I would like to fix this and understand what is going on.