4

Are Latin verbs of motion satellite-framed, verb-framed, both, or neither?

Native English verbs of motion are said to be satellite-framed: the verb usually indicates the manner of motion and a "satellite" or particle word indicates the direction. For example, "He walked into the room," "She ran out of the room," "They floated into the room." The preposition need not take an object: "He walked in", "They floated out."

Verbs of motion in Romance languages are said to be verb-framed: the verb usually indicates the direction of motion and some complementary word, typically a gerund or participle, indicates the manner. For example, "Entró caminando en la sala," "Salió corriendo de la sala," "Entraron flotando en la sala." Native speakers of Romance languages often find the English use of prepositions without an object ("particles") jarring, as in "She walked out."

I expected Latin verbs of motion to work like those of the Romance languages, but so far in my occasional studies I haven't seen it. In fact, I've come across no shortage of verbs with a prefix indicating direction stuck on a verb indicating manner: incurro, emergo, etc. This pattern extends far beyond verbs of motion, of course. And I think one normally would say Daedalus cum filio e labyrintho evolavit, not …exiit volans.

On the other hand, in Latin I've also come across constructs like respondit dicens … where in English we would just say "He answered …". And Latin seems to make much heavier use of participles where English would more likely make another construct, such as an adverbial phrase, a relative clause, or a separate sentence, e.g. Nonne rex Minos Theseum cum Ariadna fugientem persecutus est? vs. "When Theseus and Ariadna fled, didn't King Minos chase them?"


The Latin examples above are from Lingua Latina Per Se Illustrata by Hans Ørberg.

  • I think I remember (but I might be making this up) that prefixes (as in incurro) were fossil forms of former satellite prepositions that weren't otherwise present in Latin at all. – Rafael Jul 8 at 18:10
  • 1
    Yes, Rafael, what you've said makes sense (except the claim that the prefixes are to be understood as "fossil(ized)" (?) forms. The prefix in- in incurro is not a "fossil", compared to se- in secerno, which it is). To put in the terms below, it is interesting to work out whether so-called "weak satellite-framed languages" (where Path/Result is prefixed onto the verb) are derived from so-called "strong satellite-framed languages" (where it is not always required that Path/Result be morphologically dependent on the verb). – Mitomino Jul 11 at 2:35
  • 1
    Yes, Ben, you're absolutely right. Those of us who are native speakers of Romance languages find Germanic verb-particle constructions quite difficult to acquire. For those of us who learned English quite late it is nearly impossible to produce something like John danced in (instead of "Romanglish" John entered dancing. John {went/came} in dancing is also easier for us than the Manner conflation pattern involved in John danced in). – Mitomino Jul 11 at 2:53
  • 1
    Ok, Rafael, but notice that in Talmy's typological distinction, prefixes or preverbs are satellites (that's why Latin and Slavic languages are classified as satellite-framed languages). So Engl. "to ran in" and Lat. incurrere both count as satellite-framed constructions (compared to "to enter running", which is a verb-framed construction). NB: English has both the Germanic pattern ("to ran in") and the Romance one ("to enter running"). In spite of this, Germanic speakers find it more natural John ran into the room than John entered the room running, which sounds a bit more stilted. – Mitomino Jul 11 at 12:45
  • 1
    Just a clarification (incorporated into my answer below). What is important for a language/construction to be classified as "satellite-framed" is the fact that Path (i.e., directionality) is not conflated with the verb root (e.g., in the Romance verb "entrar" Path is conflated with the verbal root: you cannot distinguish Path and Motion as two morphemes). The notion of "satellite" involves lack of "conflation of Path with a verbal root". So Path, in satellite-framed languages, can be a preverb/prefix (e.g., Lat. incurrere) or a particle (e.g., Eng. ran in). – Mitomino Jul 11 at 12:48
3

Here is a descriptive answer (see below for a long, more specialized answer):

So-called verb-framed languages are "enter running" languages and so-called satellite-framed ones are "{run in/in-run}" languages (NB: for the distinction between "run in" languages (e.g., English) and "in-run" languages (e.g., Latin or Russian), see below).

What is important for a language to be classified as "satellite-framed" is the fact that Path (i.e., directionality) is not conflated with the verb root (e.g., in the Romance verb "entrar" Path is conflated with the verbal root, i.e., you cannot distinguish Path and Motion as two differentiated morphemes in "entrar"). The notion of "satellite" involves lack of conflation of Path with a verbal root. So Path, in satellite-framed languages, can be a preverb (e.g., Lat. incurrere) or a particle (e.g., Eng. run in).

So watch out for the admittedly confusing Talmy's terminology: Latin is not a verb-framed language in spite of having directional preverbs attached to the verb. As noted above, Latin is satellite-framed because Path/directionality is not conflated with the verbal root.


In the specialized literature (see below) Early and Classical Latin have been classified within the set of satellite-framed languages (see below for the transitory status of Late Latin, i.e., towards the verb-framed type of Romance languages).

Here is just a (personal) selection of some references on the topic in case you're interested to take a look at:

Acedo-Matellán, Víctor (2016). The Morphosyntax of Transitions. A Case Study in Latin and Other Languages. Oxford & New York: Oxford University Press. https://www.oxfordscholarship.com/view/10.1093/acprof:oso/9780198733287.001.0001/acprof-9780198733287

In particular, take a look at chapters 4 ('Latin as a satellite-framed language') and 5 ('Weak satellite-framed languages'). Latin and Slavic languages are classified within the set of "weak satellite-framed languages": i.e., as languages where Path/Result is prefixed (e.g., cf. Lat. Serpentes putamina extussiunt and Engl. The snakes cough the egg shells out). Latin and Slavic languages have a lot of prefixed motion verbs (e.g., Lat. incurro, decurro, excurro, percurro, etc) but, interestingly, both lack complex resultative constructions of the English kind like to wipe clean, to hammer flat, etc. E.g., see my previous post: Why can’t we wipe the slate clean in Latin? . Similarly, weak satellite-framed languages like Latin and Slavic languages tend to be different from strong satellite-framed languages like English in that the former tend to lack examples involving a complex series/accumulation of path PPs with only one verb like Tolkien's example "He still wandered on, out of the little high valley, over its edge, and down the slopes beyond" (e.g., see my previous post: How complex a motion event can be in Classical Latin ). NB: Acedo-Matellán's (2016) OUP book is a revised version of his (2010) Phd dissertation, which is downloadable at http://filcat.uab.cat/clt/publicacions/tesis/pdf/AcedoMatellan2010PhDDissertation.pdf .

Acedo-Matellán, Víctor & Jaume Mateu (2013). “Satellite-framed Latin vs. verb-framed Romance: A syntactic approach”. Probus. International Journal of Romance Linguistics 25: 227-265. https://www.degruyter.com/view/j/prbs.2013.25.issue-2/probus-2013-0008/probus-2013-0008.xml

Stolova, Natalya (2006). “From satellite-framed Latin to verb-framed Romance: Late Latin as an intermediate stage”. In R. Wright (ed.). Latin vulgaire, latin Tardif VIII, Hil-desheim: G. Olms. 253-262.

Stolova, Natalya (2015). Cognitive Linguistics and Lexical Change: Motion Verbs from Latin to Romance. Amsterdam & Philadelphia: John Benjamins.

As you can see, the satellite- vs. verb-framed distinction, originally formulated by the cognitivist linguist Leonard Talmy, is interesting for linguists because it provides a nice example for distinguishing "genetic" from "typological" classification. It is also interesting to point out that Ancient Greek has been claimed to be "satellite-framed" but Modern Greek is "verb-framed" (e.g., see https://books.google.es/books?id=RRowDwAAQBAJ&pg=PA113&lpg=PA113&dq=Papafragou+et+al.+(2006)+for+Modern+Greek,&source=bl&ots=an3LWb-r1h&sig=ACfU3U28iDrYNqRubAD9RHLocrWzUxvgwA&hl=ca&sa=X&ved=2ahUKEwiK7PO53qvjAhUQa8AKHa-CBvEQ6AEwAXoECAkQAQ#v=onepage&q=Papafragou%20et%20al.%20(2006)%20for%20Modern%20Greek%2C&f=false ).

Conversely, Old Chinese was verb-framed but contemporary Mandarin Chinese is satellite-framed (e.g., see chapter 2 of the following work: https://repositorio.uam.es/bitstream/handle/10486/661950/fan_sheng_yang.pdf;sequence=1 ).

  • Thanks: this answer really settles it! Tangentially, learning toward a topic better for linguistics.SE, it also explains to me why Russians I've encountered are also puzzled by the directional particles in English even as Russian has directional prefixes on verbs like Latin. Bottom line, I won't be reticent about framing motion the English way in Latin—but will invoke a new verb for each "segment". – Ben Kovitz Jul 12 at 18:55
  • Just to be sure: does "PP" above mean "prepositional phrase"? – Ben Kovitz Jul 12 at 18:58
  • Yes, PP means "prepositional phrase". As for your comment above ("I will invoke a new verb for each <Path/directional> segment"), notice that this is the typical case of Romance languages (latin.stackexchange.com/questions/8821/…). However, in this sense Latin is not like Romance but is not like Germanic either. This is expected since Latin, as a "weak satellite-framed language" is to be placed in the middle with respect to strong satellite-framed languages like English and verb-framed languages like Spanish. – Mitomino Jul 12 at 19:57
  • So my prediction/hypothesis is that Latin is like Polish in this respect (e.g., see the summary of the paper jbe-platform.com/content/journals/10.1075/rcl.14.1.08lew ), which seems to be confirmed if the authors of the following study on Latin are correct: mmm.lis.upatras.gr/index.php/mmm/article/view/2726/2989 – Mitomino Jul 12 at 19:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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