5

After reading Luchonachos’ previous post, whose Latin text contains an adjectival resultative predicate (claudus effectus est ‘he became lame’), the following question came to my mind:

Why is it the case that in Latin adjectival resultative constructions are (typically/basically) reduced to the ones we can find in Romance languages (e.g., Sp. Se quedó cojó ‘He went lame’; Dejó la silla vacía ‘He left the seat empty’, etc), the ones whose verb crucially lacks a manner component?

E.g., cf. Omnes consulares (…) partem istam subselliorum nudam atque inanem reliquerunt (Cic. Cat. 1, 7).

That is to say, why is it the case that Latin (consistently/systematically?) lacks adjectival resultative constructions like the complex ones typically found in Germanic languages, where the verb has a strong manner component? E.g., cf. He pushed the door open; He hammered the metal flat; He drank the teapot empty; He danced himself tired; The joggers ran the pavement thin; He shot the President dead, etc.

Probably, something similar happens with prepositional resultative phrases of the following kind: e.g., Cicero wrote his hands to the bone.

Joonas Ilmavirta
  • 113,294
  • 21
  • 192
  • 587
Mitomino
  • 8,791
  • 1
  • 16
  • 29
  • 3
    I'm not sure there's an answer to this other than that resultative constructions of the English type are rather unusual cross-linguistically -- there's no particular reason why we should expect Latin to have them. – TKR Jan 02 '19 at 01:27
  • Well, just by looking at the wikipedia entry for "Resultative", one can find examples of resultative constructions of the "wipe-clean" type in languages that are quite typologically different like German, Chinese, and Japanese... So what do you mean by "rather unusual cross-linguistically"? – Mitomino Jan 02 '19 at 01:36
  • 2
    Maybe "rather unusual" is a slight overstatement (I'm not sure the Chinese and Japanese examples are of the same type as the English), but I don't think such constructions are common enough that we should expect a given language to have them a priori. It's an interesting question, but like many "Why does language X have feature Y" questions it may not be answerable. – TKR Jan 02 '19 at 01:48
  • 1
    Of course, you're right: we should not expect a given language to have X a priori, where X can be a resultative construction of the "wipe-clean" type, an Ablative Absolute construction, a serial verb construction, whatever. But this is not the point. Notice that the question I raised has already been raised for other languages in works like the following one, which is downloadable: http://www.lingref.com/cpp/wccfl/27/paper1854.pdf Just read the first paragraph and you’ll see that these questions are typically formulated in linguistics. – Mitomino Jan 02 '19 at 02:00
  • 1
    That gets into epistemological questions about what one means by explanation in linguistics. Personally I feel doubtful about positing unobservable parameters and tend to agree with Haspelmath that "any linguist who asks 'why?' is ipso facto a historical linguist", but this is probably going too far afield. – TKR Jan 02 '19 at 02:08
  • 2
    That's right. Linguistic variation is addressed differently by functional linguists like Haspelmath and by generative linguists like Chomsky (NB: the authors of the paper I mentioned above are also generativist). But do you know what? The important linguistic difference separating "wipe-clean/float-into" languages (e.g., English) from "clean-wiping/enter-floating" languages (e.g. Spanish) was put forward by a COGNITIVE linguist: Leonard Talmy, who is an anti-Chomskian linguist! – Mitomino Jan 02 '19 at 02:16
  • I'm aware of that and am very much a fan of Len Talmy's work. But AFAIK he never tried to deal with the "why" question (do correct me if I'm wrong). – TKR Jan 02 '19 at 02:19
  • Oh, I'm very glad to see that you are a fan of Talmy's work. So that's really great! So I can talk a bit more technically to you. According to Talmy, English and Latin are precisely classified within the very SAME typological class (they are "satellite-framed" languages), whereas Spanish and other Romance languages are "verb-framed". So Latin is expected to behave like English, contrary to fact in the domain of resultatives. So notice that the question I formulated in this post acquires much more sense. – Mitomino Jan 02 '19 at 02:30
  • 3
    Is Latin really "satellite-framed"? (I've always found Talmy's terminology unfortunately opaque, BTW; "path-centric" and "manner-centric" seem clearer.) It seems unlikely insofar as none of its descendants are, nor to the best of my knowledge are other early IE languages (at least, I've looked into this a bit in Greek and found very little in the way of satellite framing). But anyway why should we expect satellite framing to correlate with the existence of English-type resultatives? – TKR Jan 02 '19 at 02:37
  • 1
    Yes, Talmy (2000: 104) classified Latin within the satellite-framed type (Talmy, L. 2000 Toward a cognitive semantics, MIT Press). As for Greek, Ancient Greek has been claimed to be a satellite-framed language, whereas Modern Greek has been claimed to be more verb-framed-like. As for the correlation between the “float-into” (path of motion) and “wipe-clean” (resultatives), it was established by Talmy himself. And, yes, other linguists have precisely criticized his correlation (e.g., in the paper I mentioned above the generative authors criticize this correlation put forward by Talmy). – Mitomino Jan 02 '19 at 02:58
  • Do you know of any specific studies of these lexicalization patterns in Latin or Greek? That would be interesting to look at. – TKR Jan 02 '19 at 03:05
  • Oh, yes. As for (Ancient & Modern) Greek, take a look at https://www.researchgate.net/publication/223009463_Grammaticalized_aspect_and_spatio-temporal_culmination – Mitomino Jan 02 '19 at 03:10
  • And for Latin, see Acedo's doctoral dissertation published by OUP: https://global.oup.com/academic/product/the-morphosyntax-of-transitions-9780198733287?cc=es&lang=en& – Mitomino Jan 02 '19 at 03:12
  • Relevant: https://latin.stackexchange.com/questions/17716/auster-misceat-mare-aestum – cmw Feb 10 '22 at 14:11
  • @cmw I understand this question could be relevant if the sentence at issue there (Si mare uoluens turbidus Auster misceat aestum) would mean what the OP intends ('If the turbulent South Wind stirs the rolling sea into a passion') but I doubt that this example can be said to contain a resultative construction. Cf. https://books.google.es/books?id=StLoix_FWTQC&pg=PA42&lpg=PA42&dq=Turbidus+Auster+Misceat+aestum&source=bl&ots=G3qP1V6BPR&sig=ACfU3U2vVIFKPH-93MczeyGTHtV_tIxClQ&hl=ca&sa=X&ved=2ahUKEwjlyvD0gvb1AhWQ7KQKHaU_AaQ4HhDoAXoECAIQAw#v=onepage&q=Turbidus%20Auster%20Misceat%20aestum&f=false – Mitomino Feb 10 '22 at 20:59
  • @Mitomino Unfortunately, your link doesn't work for me. The perils of using Google Books. I actually assumed it was an example of prolepsis, but I guess I'll wait until I get information on the link you provided before digging deeper into it. – cmw Feb 10 '22 at 22:01
  • @cmw The translation in this link (in Dronke (1994) Verse with Prose from Petronius to Dante, HUP) confirmed my intuition that this example is not to be analyzed as a resultative construction: "If the turbid South wind / whirling the sea / stirs up the breakers / ..." but perhaps you'll convince me that this is an example of resultative disguised under the guise of "prolepsis". In any case, a good answer to my question would be to show some parallelisms but also some differences between Germanic resultatives and cases of prolepsis in Latin (e.g. cf. John hammered the metal flat(*tened)). – Mitomino Feb 11 '22 at 02:38
  • I don't think I could properly answer the question as formulated, as I'm with TKR that "why" questions are ultimately unanswerable and in fact counter-productive. There is no why, there is only how. Language X creates results in A way, language Y creates results in B way. Gray areas are interesting, but they don't really answer "why." – cmw Feb 11 '22 at 05:20

1 Answers1

2

A comment of yours on another question led me to this interesting question and to an embryonic hypothesis inspired by reading a paper on "Aspect and Assertion in Mandarin Chinese" that discussed how Mandarin and English have different treatment of "2-phase verbs."

My hypothesis is that for Latin to create most 2-stage expressions it must either (1) use a prefix drawn from a limited set and add it to a verb with already inherent 2-stage semantics or (2) use two completely different verbs in some sort of subordinate relationship. To add a manner component would in effect be creating a 3-stage verb, and this Latin--and Spanish--cannot easily do.

In English, a verb rich in substantive "manner" content (a 1-stage verb) can be turned into a 3-phase expression by adding an adverb of result at the end, mimicking in an iconic way an action leading to a result.

Section 4.2 on pages 747-749 of the paper on Chinese has a good discussion of what these phase expressions are. In essence, the claim is that all languages have 0-stage expressions that express a permanent situation (e.g., "The number 3 is an odd number" is always true), 1-stage expressions that express a situation that should hold for only a limited time (e.g., "She is sleeping" implies another period of time when she is not sleeping), and 2-stage expressions that require a change from one state to another (e.g., "John arrived" requires that John was not here and then that John was here). Two-stage expressions can always be created by using two different verbs, but they can also be lexicalized in one verb, such as "arrive."

Chinese has a very productive means of creating 2-stage expressions by adding widely used co-verb suffixes from a limited set that basically translate as "on," "down," "up," "out," "into," etc. These are roughly as common as the similar English phrasal verbs, like "go on," "go down," "go up," etc. This structure, as in English, is also usable for an extremely wide variety of verbs and adjectives. Crucially, since the two stages are lexically separate, the first stage can be rich in manner content and thus add a 3rd stage. You can "wipe something clean" in Chinese (擦干净 ca ganjing) just as you can in English. Chinese makes extensive use of verbs in series for other purposes, so this use of resultative co-verbs is very natural and just as iconic as in English or even more so.

In the case of Latin and Spanish, there is no ready means to add suffixes to create similar 3-stage expressions. You have to use prefixes and add them to what are already 1- or 2- stage verbs to change them to modify their semantics (e.g., ex + eo = exeo in Latin). Since this is a limited set of prefixes that cannot readily be extended to adjectives or verbs, as in Chinese and English, and since the main verb must already have 2-stage semantics, other complex action-result combination have to be expressed by using two verbs. Also, the addition of the prefix does not add an extra stage.

A crucial difference between Chinese and English, according to the paper and my personal feeling, is that Chinese can only add aspect morphemes to the result phase; whereas English can only add them to the action phase of 2- or 3-stage expressions. In English, you can say "he is going into the store," but in Chinese you cannot use such an expression (*ta zai qu shangdianli *他在去商店里 "he is going to the store"), even though it has a progressive expression very similar in feel to English (i.e., ta zai shuo 他在说 "he is speaking").

On the other hand, Chinese can add morphemes to the result phase, since they are technically verbs or adjectives derived from verbs; whereas English cannot, because it uses adverbs or adjectives similar to adverbs.

The result is that both Chinese and English readily create the type of 3-stage expressions represented by "wipe clean," but have a different treatment of them with respect to aspect and tense modification. For Chinese, they are just two verbs used in series, allowing for a rich possibility of collocations. For English, many possibilities are also possible because the first stage is unlimited and the second stage can be any verb or adjective compatible with the result semantics of a predicate. Any manner verb plus a resultative expression gives you a 3-stage expression including manner and result.

Latin and Spanish seem to have the same types of restrictions similar to those of Chinese in not being able to add tense and aspect morphemes to the first stage of 2-stage expression. Even though Spanish has a progressive construction quite like the English and Chinese ones (i.e. él está hablando "he is speaking"), as in Chinese, you cannot say: *"él está yendo" (except in very narrow circumstances to express a habitual event). You cannot use this construction as the normal equivalent of "he is on his way somewhere."

In Latin and Spanish, you can't easily get something wiped and get it clean, so you express the three stages by saying get it clean by wiping. This is the only way to make a 3-stage expression.

That leaves Latin with only bare 2-stage verbs without a manner component. So in Latin, you cannot wipe a seat clean, but you can leave a seat empty (subsellium vacuefacio)(or in Spanish, dejar la silla vacía). To add a manner component, you must add a separate manner expression.

Vegawatcher
  • 2,700
  • 4
  • 18
  • 1
    +1! Thanks for this account based on Klein et al. (2001). As for Vmanner-Vresult compounds in Chinese, I recommend you Huang's (2006) work (https://dash.harvard.edu/bitstream/handle/1/3353765/Huang_Resultatives.pdf) and, more particularly, Fan's (2014): see chap. 4 of https://repositorio.uam.es/bitstream/handle/10486/661950/fan_sheng_yang.pdf;sequence=1 Fan (2014) & Acedo-Matellán (2010) (http://diposit.ub.edu/dspace/bitstream/2445/42060/1/VAM_PhD_THESIS.pdf ) deal with Talmy's typology of motion events, which I consider fundamental to solve why Latin lacks complex resultative constructions. – Mitomino Feb 03 '22 at 02:21
  • Thanks for the references. I am only a "self-taught linguistic" through exposure to a wide variety of languages and language families and so reading some of these works may be slow going with all the jargon and assumed knowledge of other linguistic frameworks. I'll try to spend some time digesting them, since I am quite interested in the Latin, Chinese, and Spanish. – Vegawatcher Feb 03 '22 at 21:27
  • 1
    These three languages are very useful when studying Talmy's famous typology of lexicalization patterns (https://web.stanford.edu/~bclevin/lexpat15.pdf ): Latin is a weak satellite-framed language (see Acedo-Matellán's work above), Chinese is a strong satellite-framed one (see Fan's work above; see also https://benjamins.com/catalog/cogls.3.2.03fon), and Spanish is a verb-framed one (NB: Spanish is taken by Talmy as a prototypical example of this class). These three languages offer different semantic construals of space & aspect. So you made a very good choice! – Mitomino Feb 03 '22 at 22:04
  • I really enjoyed the insights of your answer, just like I enjoyed Mitomino's postings on this topic; I have only one question: vacuafacio ??)) – Unbrutal_Russian Feb 04 '22 at 05:34
  • I assume you are pointing out that vacuafacio doesn’t exist. I meant vacuefacio with an “e.” My typing and editing is not always first rate, and I am only just now beginning to master typing in Chinese and typing Latin with the occasional macron or apex without having to go through Māori. I’ll edit the typo accordingly. – Vegawatcher Feb 04 '22 at 10:11
  • Even without the typo that this verb exists is news to me - I didn't think to search for that form. Having never seen it, I thought such a verb could only mean "to render void of contents" - you seem to have been trying to translate "to leave empty", whose Latin in the OP is given as inānem relinquere. I also thought that vacāre could be used transitively (it cant) so was confused by your choice of verb. However I now see that it's apparently found in the same Cicero passage referring to subsellia - even though there's a chance it's a corruption and he originally wrote vacua facta. – Unbrutal_Russian Feb 04 '22 at 12:41
  • I think I've figured out why subsellia vacuēfacere sounds bad to me - these verbs are all terminative, telic causatives, or the Achievement Aktionsart, generally existing in relation to stative verbs (in this case vacāre). Thus vacuēfacere "to bring to the state of being empty" like fervēfacere "to bring to a state of boiling". This necessarily involves a voluntary Agent role. Therefore subsellia vacuēfacta means "emptied, vacated etc. seats" like domus vacuēfacta "a vacated house" - the empty state of the house is the terminative goal. – Unbrutal_Russian Feb 07 '22 at 08:31
  • What makes the reading vacuēfacta possible, and even preferrable is the fact that an agent is actually present, though a metaphorical one adventū tuō, so there's a sort of anthropomorphism going on - "your coming has emptied these seats". With vacua facta it would have to be a very awkward ablative absolute. – Unbrutal_Russian Feb 07 '22 at 08:47