I agree with Lucy up to a point. The point being where the plural is or is not firmly established in English usage. It's also important to distinguish between technical and non-technical usage. I think I would use "formulae" if I was speaking about chemistry or mathematics (something I fortunately only have to do on very rare occasions) but would use "formulas" in the ordinary loose sense of the word. I refuse to use "octopodes" ("Octopi", as a professor of classics said, "shows an ignorance of three languages."). I say "hippopotamuses" rather than "hippopotami" and "hippos" more often than either. Once a word has been thoroughly anglicised, I see no reason why its inflections should not be anglicised as well. And anyway, why should Latin and Greek be so uniquely priviliged? I don't suppose anyone says "sarong-sarong", using the Indonesian/Malay plural formation rather than "sarongs".
This is not a cry of down with the classics. I studied Greek and Latin at school and thoroughly enjoyed them. I think it would be a great pity if our classical heritage were lost, but there are far more important reasons why that is so than defending the ramparts against "The data shows" or "syllabuses".
Bingley