I was wondering how so-called "middle constructions" like the English ones exemplified in (1), which are typically translated with a reflexive verb in Romance languages (e.g., see the Catalan examples in (2)), are expressed in Latin. See also this video for a very basic introduction to so-called "middle constructions" (in the sense of (some) recent linguistic literature).

(1) a. Whales frighten easily.

b. These dishes break easily.

c. Messages transmit rapidly.

(2) a. Les balenes s'espanten fàcilment.

b. Aquests plats es trenquen fàcilment.

c. Els missatges es transmeten ràpidament.

As pointed out by Levin (1993: 5), only causative verbs of change (e.g., frighten, break, open, dry, etc.) typically enter into the so-called "middle construction" (in the sense of the abovementioned links). For example, according to her, stative verbs (e.g., see, believe, etc.) do not typically enter into this construction (e.g., *Whales see easily is ill-formed).

I was wondering (i) whether a similar set of "middle constructions" can be identified in Latin and (ii) whether similar semantic restrictions like the ones pointed out above by Levin apply to their equivalent Latin constructions (e.g., reflexive and/or mediopassive constructions, constructions with deverbal adjectives ending with -bilis suffix, etc.).

NB I: here I'm specifically interested in those "middle constructions" that receive this (appropriate or not!) label in the (recent) linguistic literature above (e.g., see the links above). I.e., although it is related to this topic, my question is not whether there is a "middle voice" in Latin.

NB II: so-called "middle constructions" (again defined in the sense of links above) are typically separated from so-called "anticausative/ergative constructions" (see this link): for example, the former can be formed from agentive verbs, whereas the latter can't (e.g., cf. the well-formedness of These horses saddle easily vs. the ill-formedness of *These horses saddled easily. NB: the verb saddle (meaning 'to provide with a saddle') is necessarily agentive).

  • 1
    Perhaps of interest: thepatrologist.com/wp-content/uploads/2019/01/… Dec 13, 2019 at 15:10
  • @EthanBierlein Thanks for this interesting document. I've taken a look at it and I've realized that it basically deals with the connection between (historical) "middle voice" and Latin deponent verbs. However, I've been unable to see any reference to the treatment of so-called "middle constructions" in the sense of the links/video above.
    – Mitomino
    Dec 13, 2019 at 17:52
  • The basic answer is that Latin had already lost the middle voice, if it ever had any. Are you interested in the various substitutions under which the middle voice could be subsurmized, or only in the specific examples. I mean, I thought Spanish "the vase broke itself" was a proper continuation of middle voice, which McWorther talking about reflexive pronouns has compared to "the car broke down on me"
    – vectory
    Dec 14, 2019 at 22:15
  • @vectory As noted above, so-called "middle constructions" (e.g., These dishes break easily) are different from ergatives (e.g., These dishes broke). The former have a generic meaning and often require an adverbial expression (e.g., easily). As noted above, agentive verbs can enter into the middle construction but not in the ergative one: cf. the well-formation of These books {read/sell} well vs. the ill-formation of *These books {read/sold}. Note that "the vase broke (by itself)" (or its Romance equivalent with reflexive) is not a middle construction (in the sense above).
    – Mitomino
    Dec 15, 2019 at 17:23
  • 1
    @vectory You're right. In some cases it is not always easy to distinguish them formally. Some nice contrasts given in the first link above are: cf. *The glass broke with a hammer (ergative) with This glass breaks easily with a hammer (middle). The former is ungrammatical, whereas the latter is ok. This contrast has to do with the fact that ergatives are incompatible with instrumentals, which require an agent, whereas middles are compatible with them. That's why middles are compatible with agentive verbs (These books read/sell easily), whereas ergatives aren't (*These books read/sold).
    – Mitomino
    Dec 16, 2019 at 19:16

1 Answer 1


I will make the argument that maturo is such a verb. (which, in some respects, matches the English ripen/grow in this (which is not shared by other languages)).

First, its primary meaning is the transitive ripen: to bring to maturaty. "maturo uvas": I ripen grapes.

Now, In order to say "to ripen" intransitively (i.e., to mean "to become ripe or soft, to come to maturity") as in "honeydew melons ripen slowly" the passive verb of maturo can be used as L&S demonstrate. Yet, however, in L&S under II.A, we can find an intransitive use of maturo which seems to qualify into the definition. The only 2 examples given by L&S contain adverbs (sero and tardius):

ficūs, quae sero maturant

Another example I could locate is from Pliny N.H:

Alia etiamnum aetatis differentia. amygdala enim et pirus in senecta fertilissimae ut et glandiferae et quoddam genus ficorum, ceterae in iuventa tardiusque maturantes (... the rest, in youth, ripen slowly.

As maturo is a "a verb of change", and the examples given above are general in their nature, IMO it fits the definition nicely. Also I find the fact that maturo is used passively to convey the English intransitive meaning of ripen, to give more space, at least in theory, for a general meaning of a middle construction.

Not sure I understand the question well. On one side it seems to be interested in how to express middle construction in Latin, but also interested to identify "middle construction". I view those as 2 separated things. Here I address only one issue (namely the existence of middle constructions in Latin).

  • Your examples exemplify the so-called "inchoative"/intr. variant associated to transitive causative verbs of change of state. Cf. the links in my question above and this footnote: books.google.es/…
    – Mitomino
    May 11, 2021 at 15:20

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.