When I'm reading macronized texts, the prefix "pro" always seems to be marked long, with the exception of a few words in which it's followed by the letter "f": profugus, for example, and proficīscī, but not prōficere or prōferre. I've seen a handful of others that aren't coming to mind at the moment. Harm Pinkster's 2003 Latin-Dutch dictionary, which I believe corrects a lot of the vowel-length errors in earlier dictionaries (I'm looking at you, Lewis & Short), lists the "o"s in profugus and proficiscī as short. Ernout-Meillet isn't any help on this question.

Does this shortening follow a rule of which I'm unaware, or does it just, you know, happen sometimes?

1 Answer 1


There appears to be a significant amount of variability in the length of the vowel in pro-, and the short variant seems to originate at least in part from some common ancestor of Latin and Greek, rather than just from a process of shortening that applied in the history of Latin itself. I would guess that analogy may have played a role in the eventual distribution of the variant forms, though.

Hale and Buck's Latin Grammar (1903) says

Prō has a short vowel before another vowel or h, and before f followed by a vowel, except in prō-ferō and prō-ficiō. So pro-avus, pro-inde, pro-hibeō, pro-fugiō, pro-fundō, etc. But before vowels prōd- is commonly used; e.g. prōd-eō, prōd-esse, prōd-igo.

a. The form with the short vowel appears also in pro-cella, pro-nepōs, pro-pāgō (usually) and, in some other less obvious compounds, as pro-cul, pro-pe, pro-bus.

NOTE. Although before a vowel or h the short vowel may be the result of regular shortening (21), in the other cases pro represents an inherited variety of prō (Greek has regularly the short vowel). In Early Latin the demarcation in the use of the two forms was less fixed than later, and even in classical poetry there are occasional departures from the normal usage; e.g. pro-cūrō beside the usual prō-curō, and, vice versa, prō-fundō beside the usual pro-fundō.

(p. 10)

  • The conclusion seems to be that while there is correlation between quantity and context, one has to learn the length of pro- case by case much of the time.
    – Joonas Ilmavirta
    Commented Jun 4, 2018 at 12:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.