Caveman linguistics
« previous post | next post »
Today's SMBC:
The mouseover title: "I was trying to figure out why cavemen sometimes do and sometimes don't drop prepositions. Anyone have a theory?"
The Aftercomic:
If "cavemen" were early members of our species, and if their languages were like those of recent hunter-gatherer peoples, linguists' hypothesis about those languages would be somewhere between "just like us" and "more morpho-phonologically complex" (due to lack of the simplification caused by language mixture and second-language learning in larger groups).
If "cavemen" means "Neanderthals and Denisovans", we have no real basis to judge. But a plausible guess would be "same".
Update — Some background on pirate accents:
"R!", 11/03/2003
"Type like a pirate day", 9/9/2004
"R!?", 9/19/2005
"Type like a pirate", 9/18/2006
"Pirate R as I-R-eland", 9/20/2006
"Powarrr law", 9/20/2006
"Post like a pirate", 9/19/2007
"R", 9/9/2008
"It be that day again", 9/19./2008
"Said the Pirate King, Aaarrrf", 9/27/2010
"R R R", 9/19/2012
"Talk like a pirate", 9/19/2017
"International Talk Like a Pirate Day, again…", 9/22/2024
KeithB said,
August 20, 2026 @ 7:10 am
Ridley Scott is making a new version of "Treasure Island" with Hugh Jackman* I wonder if he is going to copy Robert Newton's accent from 1950, which I understand started it all.
*Or as they say in "Night at the Museum" Huge Ackman.
Jkw said,
August 20, 2026 @ 7:54 am
I like the inclusion of "Me not mark plurals", a sentence which is marking a plural. It is hard to actually use non-native grammar. I think he meant to follow the rules he stated, but put in 3 marked plurals and I think none that were unmarked.
Chris Button said,
August 20, 2026 @ 9:07 am
I don't think they necessarily simplify over time. They just change over time (simpler in some ways, more complex in others)
NW said,
August 20, 2026 @ 10:31 am
contain obvious mistake. caveman not know modern linguistic terminology. infinitival "to" modern analysis. caveman think "to" part of verb.
Michael Watts said,
August 20, 2026 @ 4:36 pm
It normal grammar. Contain features present in Mandarin, Russian, Latin.
Only one pronoun for the self: current in Mandarin. (I believe this is an innovation over classical Chinese.) Not a feature of Russian or Latin.
No articles: current in Mandarin, Russian, and Latin.
No plurals: mostly current in Mandarin. Some plurals are marked. Not a feature of Russian or Latin.
No copula: not a feature of any of the three languages listed. In each of those languages, it may be omitted from certain sentences that would require it in English, but none of them lack it.
No subject-verb agreement: current in Mandarin. Not a feature of Russian or Latin.
No infinitival 'to': this is hard to parse. Russian and Latin explicitly mark infinitive verbs. They don't use English particles to do it. Mandarin doesn't mark infinitival forms.
I don't understand the list of examples. Mostly these are features of Mandarin.
—
I asked a Chinese friend how she chose between singular and plural when including English nouns in an otherwise Mandarin sentence; she said she always uses singular. For me the difference is more deeply ingrained and I prefer to use the form of the English word that I would have used if the whole sentence were English.
I've tried dropping some inflection from English words while communicating in Chinese, and I don't think I dropped the right stuff.
Tom said,
August 20, 2026 @ 6:07 pm
No one will do better than the 1990 Treasure Island with Charlton Heston and Christian Bale.
Tom said,
August 20, 2026 @ 6:14 pm
That's interesting. It's an inversion of the pattern in English in which a singular is always used when initially imported from a language without pural. But I would guess it's more natural-sounding for a Chinese speaker to drop a plural than for an English speaker to add a plural since we have singular collective nouns.
Ben said,
August 20, 2026 @ 11:09 pm
@chrisbutton
"I don't think they necessarily simplify over time. They just change over time (simpler in some ways, more complex in others)"
Peter Trudgill argues that, while language complexity cannot be determined by any single factor, a higher number of second-language learners is an excellent predictor of lower linguistic complexity. (Pidgins are the most textbook example of this, though they are often treated as an outlier rather than the lower end of a continuous complexity spectrum).
Trudgill. 2011. Sociolinguistic typology: Social determinants of linguistic complexity. Oxford.
Sampson, Geoffrey; David Gil; Peter Trudgill. 2009. Language complexity as an evolving variable. Oxford.
Julian said,
August 21, 2026 @ 1:58 am
If languages have a tendency to simplify, why did they ever start out complicated?
Michael Watts said,
August 21, 2026 @ 3:21 am
What do you mean? The theory here is that languages have a tendency to complexify.
wgj said,
August 21, 2026 @ 4:33 am
I can't believe "No gendering" is left out – that one is so basic!
Michael Watts said,
August 21, 2026 @ 6:40 am
I don't think English speakers believe that it would be difficult for cavemen to keep track of who is male and who is female. Observing the distinction in speech is difficult, for modern speakers of languages with no such distinction, but this is not widely recognized.
Jerry Packard said,
August 21, 2026 @ 9:40 am
@Julian
They do not start complicated. They acquire complexity over time. The nature of the complexity is that pieces of meaning want to concatenate, and then our task becomes ‘parsing’ them into concatinated units, which units are themselves moveable and concatenatable.
Terry K. said,
August 21, 2026 @ 12:52 pm
@Christ Button and @Julian
The claim in the original post was not that languages simplify over time. It's that specific circumstances cause languages to simplify. Circumstances that cavemen may have tended to lack.
VMartin said,
August 21, 2026 @ 1:25 pm
According to the academician Nikolai Marr, the first pronouns were those of the plural number, as people had not yet distinguished themselves from the collective. Marr also examines the origin of the reflexive pronoun -self, linking it to the concept of the "head"—a connection he demonstrates through Caucasian languages. However, one must delve deeper when considering the reflexive nature of human thought. This is aptly illustrated by Hegel and, in particular, Fichte; in my view, their reflections align quite well with Marr's insights or vice versa.
pfb said,
August 21, 2026 @ 2:55 pm
Came across another "Caveman Language" comic just today!
(New to me, not to the world–it's from 2012)
https://files.explosm.net/comics/Rob/cavemen.png
Michael Watts said,
August 21, 2026 @ 3:35 pm
"Academician"? There is no possible way to support this claim. Why would we be interested in someone making up random ideas?
Chris Button said,
August 21, 2026 @ 3:47 pm
@ Terry K
I think those specific circumstances can also cause languages to become more complex. It goes for phonology too as languages morph over time in various directions.
VMartin said,
August 21, 2026 @ 4:05 pm
@Michael W.
Yes, it was Marr’s idea—and a rather interesting one. He even claimed that the plural form of nouns emerged first, with the singular being derived from it—even though, according to modern grammar, it appears to be exactly the opposite. I would add that, in this context, one probably needs to clarify the relationship between human thought, ideas, and language. When one reads Pinker’s contemporary reflections on the subject and compares them with Kant or Hegel—and their distinctions between *Verstand* and *Vernunft*—it feels like comparing things from different galaxies.
Chas Belov said,
August 21, 2026 @ 7:09 pm
@Tom:
Maybe almost always. I'd say "We coastal folks have to watch out for tsunamis." but "Give me one each of those sushi."
Bob Ladd said,
August 22, 2026 @ 9:41 am
@ Michael Watts:
"Observing the distinction [between e.g. "he" and "she" in speech is difficult, for modern speakers of languages with no such distinction, but this is not widely recognized."
I was told long ago, by a teacher who knew him, that the early 20th century Chinese-American linguist and phonetic wizard Y. R. Chao came up with a solution to this problem for his wife, who constantly had trouble with the he/she distinction in her English. He coached her to produce a fairly weakly articulated IPA [çi], which could be used for both "he" and "she", and most listeners supposedly accepted the result as a foreign-accented version of either pronoun according to the context. I can't vouch for this story myself, but it sounds like a reasonably plausible trick.
Rodger C said,
August 22, 2026 @ 9:45 am
Marrists still walk the earth???
Jerry Packard said,
August 22, 2026 @ 11:27 am
From Packard 2021 ‘A Social View on the Chinese language’: (keep in mind that this is merely a theory, meant to be criticized and improved upon)
At present, most Chinese words consist of two syllables, but those two syllables are no longer primarily the free words that make up compounds, but are mostly either bound roots or affixes.10 So in terms of linguistic morphological typology,11 Chinese has shifted from being a fusional language (from proto-Sino-Tibetan to proto-Chinese) to an analytic, isolating language (from proto-Chinese to Old Chinese) to being an agglutinative, compounding language (Middle Chinese to Modern Mandarin), and now appears to be cycling back in the direction of becoming a fusional language, with the development of derivational and inflectional morphological processes.
Michael Watts said,
August 25, 2026 @ 3:29 am
I guess Chinese could be "cycling back in the direction of becoming a fusional language", but I'm not sure why we'd say that. There are no fusional elements currently in existence that I know of.
By my analysis, inflection in Mandarin nouns is… weak at best. Nouns can reduplicate to form a diminutive. (Syntactic reduplication is everywhere in Mandarin. In verbs it also has a kind of 'diminutive' sense. In adjectives it's emphatic. In measure words it gives the sense 'every', which may be weakened to 'many'.)
Inflection in verbs is more robust. There are *two* kinds of verbal inflection:
– Verbs may be marked with one (or none) of the aspect markers 着 (continuous), 了 (perfective) or 过 (experiential). These do not adjust their form for any reason.
– Verbs may be marked for the modality of "possible" or "impossible". In order to do this marking, the verb must include a result complement, and the dummy complement 了 (not identical with the perfective aspect marker, but spelled the same) is available in case it doesn't. For possibility, 得 is inserted between the verb and its result, and for impossibility, 不 is.
I think the case that these are both forms of inflection is strong. But it's hard to call them "fusional"; my impression was that "fusional" referred to morphemes that simultaneously convey information in several dimensions. These do exactly one thing.
Jerry Packard said,
August 25, 2026 @ 10:10 am
@Michael Watts
Thank you.
I guess the difference may be in my (mis)understanding of the term fusional. Over time my students and I have noticed that the 2nd syllable of Mandarin words has phonologically and semantically bleached to become more & more functional. I don’t have many examples at hand (I will have to go find them, including papers that my students have written) so I’ll go with the easy and obvious -le and -de. Keeping it simple (!), narrow-scope -le/-la is multi-functional, as is -de as a marker of, among other things, (sentence) embedding [这本书贵的没有人要买 this book expensive -de nobody want buy ‘this book is so expensive that nobody wants to buy it.]. So -de (的，得，地) is clearly multi-functional including its use, as you mention, as the ‘resultative’ inflectional marker.
But, as you note, most of these are phonologically invariant. My prediction would be that as the language moves from characters to pinyin, the phonology will begin to bleach out.
Chris Button said,
August 25, 2026 @ 5:39 pm
I've always loved Gordon Luce's statement on Chin languages in his 1959 article “Chin Hills – Linguistic Tour (Dec 1954) – University Project”:
“I was brought up to regard Far Eastern languages generally as (i) Monosyllabic
(consisting of words of one syllable); (ii) Invariable (not modified by any inflexions); and
(iii) Isolating (destitute of syntax). Chin is a language which disproves all three
statements.”
The complex verbal inflections really make them stand out relative to other Sino-Tibetan / Tibeto-Burman languages.
I suppose the danger is to speculate that the inflections come from a host of proto-suffixes (e.g. -t, -k, etc.) instead of looking at the specific conditioning environments that could result in the various reflexes. Such speculation reminds me a little of how Old Chinese is sometimes reconstructed with a plethora of supposed affixes despite paltry evidence for all but a very select few of them.
Michael Watts said,
August 25, 2026 @ 5:54 pm
I don't disagree with your comments, other than that I understood "fusional" differently. There is a case where I don't think we're talking about the same thing.
这本书贵的没有人要买 [prescriptively, I think the particle here should be 得; it is definitely true that native speakers don't draw any mental distinction] is not what I was referring to as a result complement, and I would hesitate to call it an example of verbal inflection. (To me it looks like a pretty close match to English "this book is so expensive that nobody will buy it", where I wouldn't call that an inflection on is.) I was referring to the result complement seen in sequences such as 看到 "see", 猜错 "guess wrong", 听懂 "hear and understand" etc. These are highly grammaticalized and, as I mentioned, if you need your verb to belong to this class when it otherwise wouldn't, the dummy complement 了 [liao] exists for that very purpose. So e.g. 看不到 "cannot see" / 看不了 "cannot look".
In this context 得 occurs between the verb and the complement to indicate that we are speaking of the action of the verb (and the attendant result) being possible, not to indicate that the action of the verb occurred, but in a certain way or with a certain result.
In my experience native speakers communicating in text do not confuse this 得-the-infix-of-possibility with 的 anywhere near as often [I don't recall ever seeing it] as they do the 得-of-introducing-a-full-postverbal-manner-clause… but I am willing to believe that this is mostly because it's easy for a software input method to choose the correct 得 in that case.
(I should harrumph here that I don't follow Victor Mair in being willing to call 贵 a verb. Consider "postverbal manner clause" a convenient shorthand.)
I'd be interested in what this means. Phonological bleaching is pretty straightforward in that you see the second syllable of words develop "tonelessness", clitic tone. Semantic bleaching, sure. A word doesn't need to be transparently composed of meaningful parts, and when it is the ordinary meaning of those parts won't necessarily have much to do with the meaning of the word.
The first thing that comes to my mind here is the large class of nouns where the second syllable is a basically meaningless 子. These seem [I am speaking without any special knowledge] to have come from a process in an earlier stage of the language where a meaningful syllable was attached to a disambiguating 子 to signify something like "the noun, [whatever the first syllable was]". But I'm not sure that process should be seen as something currently active in the language.
Anyway, though, it's not clear to me what you mean by saying that the second syllable of two-syllable words is becoming "functional". Do you mean more than that "there is a lexical entry with two syllables, and it specifies this second syllable"?
Chris Button said,
August 25, 2026 @ 6:13 pm
@ Michael Watts
I'm with Victor Mair on this one (I so wish he were still here to be able to chime in).
After all, adjectives don't really exist… Claims to the contrary contributed to my decision to avoid syntax like the plague when I was formally studying linguistics :)
Michael Watts said,
August 25, 2026 @ 8:18 pm
One other thing I should note about 这本书贵得没有人要买 is that, according to my training, the 贵 or other verb/adjective can be separated from the 得. Kind of. The clause preceding the 得 can position a verb wherever, and 得 will follow that clause, but in the case where the verb isn't at the end of the clause, it will be repeated so that it still occurs immediately before 得.
Caveats:
– This isn't a construction that really comes up in my conversations, so I'm going off textbook knowledge instead.
– By convention this is all viewed as one sentence, but I don't see why it couldn't be viewed as one sentence (with no manner clause) followed by a paratactically related sentence that is just the verb of the first sentence, followed by a manner clause.
Something like 我看他走路走得很慢 versus 我看他走路 – 走得很慢。
Michael Watts said,
August 25, 2026 @ 8:22 pm
He's addressed the question on Language Log before. Specifically, he said that he took the terminology from his wife, who liked it because English speakers studying Chinese wanted every sentence to contain a verb. At no point did he ever make a motivated argument that Mandarin adjectives should actually be considered verbs.
I mentioned that Russian, for example, also doesn't use verbs in predicative sentences of this kind, and that in Russian there is no possibility of confusing and adjective with a verb. Victor Mair was not one to make, or understand, linguistic arguments.
Jerry Packard said,
August 26, 2026 @ 8:59 am
@Michael Watt
Very cool.
By 2nd syllable bleaching/reduction I mean phonological and semantic, with phonological meaning neutral tone and coda attrition, and semantic meaning the usual bleaching and also movement rightward along the grammatically cline. So by functional I mean more like a function word than a content word.
gui4de 贵得 ‘so expensive’ was the first thing that popped into my head, but you can use just about any verb: 这个人跑得没有人比他快 this-m person run de nobody faster than him ‘this guy runs such that nobody’s faster than him’.
When you say ‘I wouldn't call that an inflection on is’ you might do so if the embedding marker/complementizer was juxtaposed/affixed to the verb as it is in Mandarin.
The result complements you refer to are easily considered inflection on left-headed 2 syllable verbs 看到 /看得到/看不到 is the same (see Packard 2000 ‘the Morphology of Chinese’ for detailed treatment). My student Junghwan Maeng wrote a class paper several years back arguing that the -dao that occurs as verb-dao is bleaching its way to becoming an inflectional affix, using corpus examples such as 跟到， 给到 and 对到.
“…degenerate’ resultative V₂ endings such as -dào ‘attain’ … are a good example of grammaticalization in modern Mandarin. Recall that in gestalt V₁–V₂ words containing these endings, the theta role and semantic information that had been ascribed to the ‘erstwhile-V₂-words-turned-degenerate-endings’ became part of the head verb V₁ rather than remaining part of the erstwhile V₂ words. As a result, these forms have evolved into endings that reflect aspectual properties of the V₁ verb (viz., perfectivity; referring to the completion of verb action) rather than reflecting the full meaning and argument structure of the erstwhile V₂ words. Thus, ‘degenerate’ endings like dào have lost some of their ‘content’ and evolved toward indicating grammatical ‘function’. “ Packard 2000 p. 262
I do believe X子 is a reasonably productive process, given the hundreds, nay thousands of forms. An even better example would be -er 儿 suffixation, which is not only wildly productive, but *always* causes phonological change on the suffixed syllable.
My student Junghwan Maeng wrote a class paper several years back arguing that the -dao that occurs as verb-dao is bleaching its way to becoming an inflectional affix, using corpus examples such as 跟到， 给到 and 对到.
@Chris Button
“Old Chinese is sometimes reconstructed with a plethora of supposed affixes despite paltry evidence for all but a very select few of them.”
Sagart, Bodman, Baxter, Schuessler and others may not agree that the evidence is paltry.
Chris Button said,
August 26, 2026 @ 11:18 am
@ Jerry Packard
I think Schuessler might agree with me. If we're lucky he might comment directly here.
Jerry Packard said,
August 26, 2026 @ 12:33 pm
“… word derivation by prefixes and suffixes constituted one of the most outstanding characteristics of Proto-Chinese.” Schüssler, Axel (1976). Affixes in Proto-Chinese. Wiesbaden: Verlag. p. 129
Chris Button said,
August 26, 2026 @ 3:13 pm
I agree with that statement. I was just saying that only a few of those affixes are well attested. I think that is hard to deny.
I also think that perhaps the plethora of poorly attested supposed suffixes may be handled by advances in our understanding of Old Chinese phonology rather than by speculative ideas about inflectional morphology.
Chris Button said,
August 26, 2026 @ 3:32 pm
For "suffixes", read "affixes".
Jerry Packard said,
August 26, 2026 @ 3:49 pm
@Chris
I think you have a healthy perspective on this issue. It is fun to speculate, though.
wgj said,
August 26, 2026 @ 10:03 pm
I can't think of an equivalent formulation of 这本书贵的没有人要买 in classical Chinese that has the same grammatical structure, which makes me suspect that the "[adjective] de [description]" construct is a foreign import into vernacular Chinese. The most likely suspect would be Mongolian (during the Yuan dynasty) – can anyone confirm or deny that such structure exists in Mongolian?
Chas Belov said,
August 27, 2026 @ 12:14 am
I too wish Victor Mair were still here.
@Michael Watts:
I can't speak for Mandarin, but the Defense Language Institute Cantonese Basic Course stated that all Cantonese adjectives are verbs. Further, it defined a word in Cantonese as anything that could stand as a sentence all by itself. Otherwise, it would be a bound-form. According to that rule, the word for cat would be a valid sentence in Cantonese (I'd guess requiring the appropriate context from a preceding sentence).
wgj said,
August 27, 2026 @ 3:51 am
Guys, I think when you say "verb", you mean "predicate". Chinese indeed allows far more types of word to serve as predicate (verbs, nouns, adjectives, and most famously, subject-predicate subclauses), but verb in Chinese is more or less the same word type as verb in Indo-European languages.
Chris Button said,
August 27, 2026 @ 6:24 am
@ wgj
I mean "verb". I don't think the class of adjectives exists.
Jerry Packard said,
August 27, 2026 @ 9:07 am
@wgj
I don’t know about a structure borrowed from Mongolian, but I guess the comment I’d make is that I’m not sure how much we can infer about what classical can tell us about the spoken language of the time. Victor could surely help us here.
Failing that, I asked AI: can you give me a classical chinese structure that expresses the construct ‘x to the extent that y happens?
AI responded:
common Classical Chinese structure for "X to the extent that Y happens" is 以至於 (yǐ zhì yú), which means "so much so that" or "reaching the point where."How the Structure WorksFormula: [Condition X] + 以至於 + [Result Y]以 (yǐ): Means "to take" or "by means of."至於 (zhì yú): Means "to arrive at" or "to reach."Combined: The action or state of X reaches the destination or degree of Y.Example UsagePhrase: 勞以至於病 (láo yǐ zhì yú bìng)Meaning: Working too hard to the extent that you fall ill.Breakdown: 勞 (overwork) + 以至於 (to the extent that) + 病 (illness/fall sick).
which strikes me as kinda ‘direct translatese ‘
wgj said,
August 27, 2026 @ 10:34 am
@Jerry: 以至於 is not classical Chinese, but vernacular – old vernacular, but vernacular nonetheless.
Jerry Packard said,
August 27, 2026 @ 12:58 pm
Ok!
David Marjanović said,
August 27, 2026 @ 3:00 pm
Russian lacks one. If you're thinking of есть, that hasn't been a copula in centuries; it means "there is" – and even "there are", because the marked plural суть is obsolete.
Member of the Soviet Academy of Sciences until the real linguist Arnold Chikobava talked Stalin out of it.
I would say it's an obviously silly idea; there's no motivation behind it other than to import Stalinism into linguistics (and into everything else, too).
Genius. (Unsurprisingly.)
Michael Watts said,
August 27, 2026 @ 7:05 pm
Well, I took the position that the infixes -得- and -不– are inflection.
And I'm certainly sympathetic to the idea that this "achievement" 到 is bleaching its way to becoming an inflectional affix. Stated in terms that weak, I agree. It is bleaching, and it is plausibly on the way to becoming an inflectional affix.
I have more trouble with the idea that all complements permitting the marking of possibility in this manner are "easily considered inflection". As I understand it, almost anything [that is short and expresses a state] might serve as such a complement, and this freedom means that the complement must have some independent existence. 到 is close to being maximally grammaticalized (and 了 has achieved it), but there is a wide spectrum of words/phrases/morphemes that use the same syntax. (For something healthily in the middle, think of 洗干净 "wash clean" or 擦干净 "scrub clean"; this is a relatively conventional result, but I really wouldn't want to postulate an inflectional system with a suffix for expressing cleanliness. I think 干净 must continue to be considered a separate word.)
In this case I cannot directly speak to Russian and was relying on information I found on the internet. https://russianlanguageguide.com/russian-verbs/be/ is representative. It is indeed talking about есть; it notes that the verb is not used in the present tense, but that it is still used in the past and future. Assuming this is accurate, I would say that "Russian lacks a copula" is untrue?
I note that the sample sentence given there, Я часто бываю в Москве "I am often in Moscow", is difficult to interpret using a sense of "there is".
Jerry Packard said,
August 28, 2026 @ 1:43 pm
@Michael Watt
“I have more trouble with the idea that all complements permitting the marking of possibility in this manner are "easily considered inflection". As I understand it, almost anything [that is short and expresses a state] might serve as such a complement, and this freedom means that the complement must have some independent existence”
…
I’m not sure I know what you’re saying, but in my conception the complementizer that marks embedding (-de) can be affixed to the verb, like :
他喝的连外国人也喝不过他 ta he-de lian waiguoren ye hebuguo ta. He drink-de even foreigner drink-not-pass him ‘He drinks such that even foreigners can’t outdrink him’ .
Here, the -de acts as a complementizer that allows the embedding of the embedded sentence [lian waiguoren ye hebuguo ta] ‘even foreigners can’t outdrink him’ into the main/matrix sentence [ta he] ‘he drink’
“擦干净 "scrub clean"; this is a relatively conventional result, but I really wouldn't want to postulate an inflectional system with a suffix for expressing cleanliness. I think 干净 must continue to be considered a separate word.”
…
There’s no conflict here. 擦干净 is fine; the inflection is the operation that suffixes -de to 擦 ca ‘scrub’ to derive 擦得干净 scrub-hde clean ‘able to scrub clean’ (scrubcleanable ） or 擦得跟爸爸一样的用力，scrub-de with papa same effort ‘scrubs it to the extent that he uses the same effort as daddy’.
In these two examples, the -de is a marker of embedding, it is a functional element that embeds a simple ‘ganjing’ in the first example, and it embeds a sentence in the second example. It is a functional marker of embedding that inflects on the head (left-hand member) of the verb ca ‘scrub’
Jonathan Smith said,
August 28, 2026 @ 2:59 pm
Pretty sure there are (at least) two separate "V得'complement'" situations in the standard language, call them (1) 'descriptive' and (2) 'abilitative' — while rarely ambiguous in practice, something like "做得好" for example could mean either 'did/does [it] well' (> 'Nice job!') OR 'can be done right/well (by agent)' (from online: "別人一個小時做得好，我要四小時才做得好".) The more involved stuff, sometimes involving 'extent' (e.g., 做得死去活來, examples by posters above, etc.), looks like it belongs to (1) 'descriptive'.
And more to the point, I don't see any advantage to considering zuo4de “做得” etc. to be an "inflection" of zuo4 做 — what (two different) paradigmatic slots is this "inflected form" realizing"? Better to see this as a linking "function word(s)" in 3-part "construction(s)", as the tradition has it.
It is of course interesting to think about the potential for Chinese etc. languages to move "rightward on the grammaticalization cline" — but there are real-looking examples of this in many non-standard Mandarin languages; see e.g. LaMarre (2015), "The morphologization of verb suffixes in Northern Chinese," with examples of incorporation of "aspect particles" producing "inflected" looking perfective, stative, etc., forms of verbs. (The author points out that "Standard Written Chinese may be the worst variety of Chinese to observe in order to shed light on the various morphological processes at work in Sinitic."
Re: history, first paragraph things can be striaghtforwardly seen as grammaticalizations involving de2 得 'achieve' — no need for a classical precedent. This is typical of the modern vernaculars… Hokkien varieties grammaticalize kau3 'get to' (reduced to e.g. > kah) in "extent"-type constructions in a parallel way. Not to say such processes can't at the same time also involve influence from linguistic neighbors!
Re: Mandarin "adjectives", the words in question can definitely do more verb type stuff than English etc. adjectives: …貴了很久，也會一直貴下去 etc. etc. But they are nonetheless distributionally different than "verbs": e.g. 很X etc. makes an "adjectives"-only predicate, the comparative constructions need "adjectives," etc. So collapsing these categories is theoretical goofing.
Chris Button said,
August 28, 2026 @ 4:14 pm
@ Jonathan Smith
My comment, at least, was meant for way more than just Mandarin. And, to your point, of course different verb types function differently.
The idea that adjectives don't exist is neither novel nor radical in the slightest. But non-linguists (or armchair linguists grounded in western linguistic theory) might take it that way.
By the way, going back to my citation of Luce earlier regarding Chin (not Chinese) languages, even the numerals there inflect like verbs!
Michael Watts said,
August 28, 2026 @ 5:18 pm
I agree with Jonathan Smith that these are two unrelated constructions and there is nothing to be gained from analyzing them together.
Yves Rehbein said,
August 29, 2026 @ 6:03 am
Infinitive (though "Mandarin doesn't mark infinitival forms", @Michael Watts) or participle?
For "to be or not to be" suggests 存在还是不存在, cúnzài 存在 "to exist, be; existence" (note that Latin existens is a participle). The phonetic radical 才 in both morphs is cognate to zǐ 子, which appears as semantic in 存, indicating a relation, though not ancient. See also 我思故我在 (Wǒ sī gù wǒ zài, "I think therefore I am"). Zài 在 "to exist" in Chengdu also has a pronounciation de4, incidently, vis-a-vis 得 de2 (Chengdu), that is fourth tone in Teochew in the sense of "get", for what it's worth (@Jonathan Smith). If 存 is from 在 with demonstrative *-n (according to Schuessler; Wiktionary), it also shows a sort of legitimate reduplication (@Michael Watts).
Yves Rehbein said,
August 29, 2026 @ 6:05 am
Curiously, this reminds me of Japanese 緑 midori, cf. cǎo 草 or perhaps sè 色, and Japanese inu, cf. gǒuzǎi 狗仔, Northern Min ě-ciě, so something about prenasalized stops (@Chris Button).
Jerry Packard said,
August 29, 2026 @ 7:25 am
“I agree with Jonathan Smith that these are two unrelated constructions and there is nothing to be gained from analyzing them together.”
A lot of Ch. linguists would agree with you.
Yves Rehbein said,
September 5, 2026 @ 11:19 pm
I reread the whole discussion on 緑 midori (LLOG 10/27/23) and if Eirikr is correct to compare Old Japanese manyogana mi₁do₂ri with [取る] to₂ru "to take", it seems that Shang 取 is nearly identical to 得 in meaning, composition and less so in reconstructed pronunciation. For composition see now Wandel and Herren, “Ear-y fish” in BASOR, and of course the auricle (!) bone inscriptions.
Yves Rehbein said,
September 6, 2026 @ 4:22 am
That's the Bulletin of SOAS, not the Bulletin of ASOR.
By examining the term for “shell” in Nakh-Dagestanian languages (e.g. Avar ччугlигlин /č:uʕiʕin/, lit. “fish ear”) and highlighting metaphorical associations of related words, e.g. German Ohrmuschel “auricle”, we argue that läksañña klautso refers to shells and conches rather than gills.
Wandl F, Herren AR. Ear-y fish: shells in two Tocharian B medical prescriptions. Bulletin of the School of Oriental and African Studies. Published online 2026:1-19. doi:10.1017/S0041977X26101219
Compare 取:
耳 (“ear”) + 又 (“hand”) – to take the ear of fallen prey, a rite of hunters mentioned in the Rites of Zhou. [1]
and 得:
The ancient form of 得 is 㝵, composed of 貝 (“cowry”) + 又 (“hand”) — to pick up a cowry > to obtain valuables. … [2]