Rhinoceri vs. Rhinoceroses: A Ban on Unstressed */əsəs/?

|

Is it two rhinoceri or two rhinoceroses? The spelling of final -os (instead of Latin -us) told me that any -i plural would be a new formation. The -os is usually a sign that the learned word is actually Greek (and rhinoceros is the compound of Greek rhino- 'nose' plus ceros 'horn').

Yet, I have to confess, I favored rhinoceri because I wasn't very phonologically happy with having to deal with the another -es (especially because the singular already has Neo-classical antepenultimate stress). And lucky me, the keyword rhinoceri pulled up tons of hits on Google, include a possible entry from the Merriam-Webster dictionary.

What's interesting to me here is that my preference suggests that English phonology may be driving the formation of some neo-classical plurals. That is, some neo-classical plurals may be the result of a grammar trying to avoid unstressed /əsəs/ sequences. It would be interesting to devise some sort of experiment to determine if there is a pronunciation preference.

FYI - Depending on how you analyze the post-Greek compound, the etymologically correct plural may actually have been rhinocerotes or possibly rhinocera (if the original Greek was rhinoceron).