Grammaticalization in Chinese: Gradual or Sudden?

They say that language change is a gradual process, like most types of change. Rome was not built in a day, and Latin did not become Italian, Spanish, Portuguese, and French in one day, either. We did not go from Chaucer's English to the English on Hubpages without countless specific changes along the way.

But what about each of those changes? Did they come about gradually, too? Or was each step along the way sudden and discrete?

Take Modern Mandarin, the Chinese spoken in both Mainland China and Taiwan. How did it get to be the way it is today? From the time of Late Archaic Chinese spoken by Confucius to the language used by millions of people in the world today, was the change gradual and continuous, or was it characterized by sudden and discrete events?

Image Credit: Wikipedia
Image Credit: Wikipedia

In his article, "Grammaticalization in Chinese", Walter Bisang considers whether  the change in usage of certain constructions from Late Archaic Chinese to Modern Mandarin could  have taken place gradually or were these developments discrete and sudden.


Grammaticalization is the historical process by which grammar is formed from non-grammatical or less grammatical elements in a language. An example in English of grammaticalization is the development of the ending -ly that can be added to nouns to make them adjectives or to adjectives to make them adverbs. Originally a noun meaning "body", the -ly ending went through a process of reduction, affixation and semantic bleaching on its way to become a part of the derivational morphology of modern English. Old English "lic" meant body. When added to the word "man" it made "man-lic", meaning "man-like". In fact, modern like is derived from a dialectal variant of "lic". Eventually, "lic" acquired a meaning more like "like" and less like "body". It began to be affixed to all sorts of words, such as craeft-lic, "craft-like" or "skillful," and it eventually acquired the more grammatical property of deriving adjectives out of nouns. At some point the final consonant was dropped, along with the stress on the last syllable, and we ended up with words like manly and friendly.(Ageo and Pyles, The History and Development of the English Language.")

The original meaning and function of lic was that of a noun to refer to the corporeal concept of "body", but by the time the grammaticalization process was done, the resulting -ly became a very abstract grammatical element whose function was to mark categoriality. It took a long time for this language change to be accomplished. It did not happen all at once.

So, in terms of the development of -ly in English, we see a gradual process in which a lexical item (the word lic meaning body) started out independent, then became attached, then lost its stress and got reduced and eventually became an affix. But does it always work that way when grammar changes? What about languages that have few if any affixes and hardly any inflection? Is the formation of grammar gradual there, too?

Grammar in Late Archaic Chinese

Late Archaic Chinese had two significant characteristics that come into play here:

  1. Categoriality was underspecified. The same word could be a noun or a verb or an adjective or almost any other part of speech, if it was put in a syntactic slot that was intended for that part of speech.
  2. There was a lot of ambiguity about what specific slots in a sentence could be used for.

For instance, in Modern Mandarin, if we want to say "You are my friend", we can say it like this:

你 是 我 的 朋 友

ni shi wo de pang you

you COP I POSS friend friend.

"You are my friend"

The two word combination of 朋友 peng you contains the Late Archaic word 友 you. In Late Archaic Chinese 友 you meant friend, too, but it could be used as a verb, or a noun or almost other part of speech. So, for instance, in the following example sentence from Bisang's article, "Grammaticalization in Chinese, 友 you functions as a verb:

Another way of rendering the sentence in English, one that emphasizes the "verbiness" of 友 you is to translate it like this: "What I am to Yan Ban is that I friend him."

This is quite similar to the usage of the English word "friend" as a verb in the context of Facebook. "Please friend me."

According to Walter Bisang, how entrenched a word is in its prototypical grammatical category as a noun in Late Archaic Chinese depends to some extent on the following hierarchy:

1st/2nd person > proper names > human > nonhuman > abstract

The further to the left it appears in the hierarchy, the less a nominal element is likely to be interpreted as a verb when placed in the verb slot. But even proper nouns can function as verbs, as seen in the following example:

Another way to render this sentence in English to show the "verbiness" of the proper name Wu King is to translate it into colloquial English like this: "Gong Ruo said: You wanna Wu King me, huh?" (Notice that the interrogative particle hu functions a lot like English "huh?")

In view of how free words in Late Achaic Chinese were to change from one category to another, the idea of gradualness in category shift does not seem to apply. Another factor to be considered was that one and the same sequence of words could be interpreted as meaning different things, if seen through the semantic overlay of different constructions. Walter Bisang gives the following example:

According to Bisang, in example (7) above, the sequence of words bing bu xing (ill not fortunate) has more than one syntactic interpretation, because there is more than one construction that we could fit the words into and because the word bing could be made to belong to different grammatical categories.

Can we do this in English? Without morphological markers, can the word ill be used for different grammatical categories? Yes, although English is a little less flexible than Chinese in this:

a. The ill are unfortunate. (Ill is a noun, and unfortunate is an adjective).

b. The unfortunate are ill. (unfortunate is a noun and ill is an adjective).

Even though English is good at employing nouns and adjectives interchangeably depending on which slot they occupy, English keeps which slot is which more distinct by employing articles (like the) and copulas (like are) to mark which is the noun and what is its complement.

However, in colloquial exchanges, we can even use the word ill to stand for a whole clause as in 7(c) above:

c. Ill? Unfortunate! (Meaning: "Is he ill? How unfortunate!")

How Can We Tell What a Word Means in an Ancient Chinese Text

In any event, what with the underspecified categoriality of words and the multitude of different ways the same slots in the exact same sequence could be interpreted, it is hard to say what exactly would qualify as a gradual change from Late Archaic Chinese to Modern Mandarin. Simply observing that a word went from being used as a noun to being an adjective or verb would not require the passage of a single day. It would just depend on what slot you stuck it into in a sentence. And since the same sequence of words might be interpreted as different constructions depending on the context, it would even be hard to say that a change in syntax occurred simply because the same sentence got a different reading.

So this brings up a very difficult question: how do linguists know when a historical change has occurred in Chinese syntax?

This question is especially interesting to me and to my friend June, because we are trying to track the development of the word 是 shi back in time to determine which usage came first, pronoun or copula? It's maddeningly difficult to tell, because as far back as we can go, 是 shi seems to be used for both in different passages in the very same ancient texts.

There are several ways in which we can deal with this problem. We can assume that one of the usages came first and the other is derived from the first one. Or we can say that 是 shi is underspecified for "nouniness" or "verbiness" in ancient Chinese. Or, we could make some kind of argument about the development of constructions.

There are a lot of pitfalls for the unsuspecting when trying to decide how to read an ancient text. We can't call upon native speakers of ancient Chinese or even Late Archaic Chinese to help us out, because they're all dead. Speakers of Modern Mandarin have their own intuitions about older texts in Chinese, but their views are often colored by the way Chinese is spoken today.

Walter Bisang writes: "Of particular for the understanding of when grammatical changes took place is the comparison of near identical passages in traditional texts that underwent redactional changes with texts that did not." (Bisang 2010.273).

What does this mean? If there is a traditional text that was recopied for purposes of teaching and "proper", formal language was used, we may learn a great deal from slight changes that occur in some passages and not in others.

Bisang is particularly interested in the development of resultative construction in Chinese, which consist of a sequence of two verbs the first of which describes an action and the second describes its result. We don't have something like this in English, but the closest we have is a verb followed by an adjective: "I shot him dead" or "I bored him silly." In Chinese what you would say would more like "I shot die him" or "I bored stupefy him" where the first verb stands for the action and the second for the result.

In the following two examples, Bisang explores the development of the resultative construction in Chinese by comparing the same passage in two versions of the same book written almost two centuries apart.

The earlier text in example (37), Bisang argues, uses a resultative construction zhan sheng "battle win", meaning "fought and won" or "fought until the end result was winning," but the later text, which is more conservative in register, used a single verb bai "defeat", because even though the resultative construction was already in existence in 195 BC, it still wasn't considered proper formal Chinese 195 years later.

This way of arguing seems a little counterintuitive, because the earlier text is considered more "progressive" linguistically than the later one.

Bisang observes: "A number of V-V constructions from late archaic Chinese and premedieval Chinese may look like resultative constructions from the perspective of Modern Standard Chinese. But such an analysis would be anachronistic." ("Grammaticalization in Chinese", p.272)

At the synchronic level, Bisang concludes that there is no continuity or gradualness between different constructions or between the use of the same word in different slots. But over time there is some continuity or gradualness in development of syntactic constructions, because the same sequence of words may be interpreted in different ways simultaneously, and because the criteria for choosing one interpretation over another are gradient.


So what about English? Is nouniness versus verbiness a graded quality? Can we measure the degree to which "friend" is a noun as compared to some other word? Or is it just that English words, like Chinese words, are underspecified for grammatical category?

Bisang offers the example of the English word "pipe" as one that can be used as both a noun and a verb. We can pipe natural gas from one location to another (verb). We can say this pipe is so many inches thick. Is the word pipe underspecified for category? Bisang thinks not: " In the case of English, it is not possible to find a general semantic rule which derives the semantics of pipe in its verbal use from its more frequent nominal use... For that reason it is necessary to assume two different lexical entries for pipe." ("Grammaticalization in Chinese" p.257)

I'm not so sure about that. It's got to be a lot easier to derive the meaning of "pipe" as a verb from its meaning as a noun than to get a verbal meaning of "to kill" out of the proper name "Wu King."

What do you think? When we started using "friend" as a verb on Facebook, was the process gradual or sudden?


Algeo, John and Thomas Pyles. The History and Development of the English Language. Boston: Wadsworth.

Bisang, Walter. "Grammaticalization in Chinese". in Traugott and Trousdale, Gradience Gradualness and Grammaticlaization. Amsterdam and Philadelpia: Benjamins.

Bisang, Walter. “Precategoriality and argument structure in Late Archaic Chinese”, in: Leino, Jaakko (ed.) Constructional approaches to language. Amsterdam & Philadelphia: Benjamins

(c) 2010 Aya Katz

Do nouns become verbs in English gradually or suddenly?

See results without voting

More by this Author

  • Project Bow's FAQ: Why is it called a lexigram when it looks like a word?

    At this point in his development, Bow no longer needs lexigrams. He is literate and he spells out his own words. Bow developed literacy at the age of five and half years. But before that, we used lexigrams. And our...

  • Language is Learned

    You may have heard it said that all babies have the same linguistic abilities at birth. This is true. They have no linguistic abilities at birth. No child is born knowing language, and early infancy and childhood are...

  • Why Prejudice is Bad

    Everybody agrees that prejudice is bad. I have never in my life met anyone who was in favor of prejudice. There is such universal agreement on this topic that it makes us feel united. Except when it comes to actual...

Comments 10 comments

CaribeM profile image

CaribeM 6 years ago

Wow, I'm very impressed by your article! Thanks for posting! Very useful for those (like me) who are trying to learn Chinese. It is a very interesting language, sometimes very precise and concrete, some other times very metaphorical and idiosyncratic.

Can you write something about cultural/historical differences, "world views" and language?

Hope to see more of your articles soon. ; )

drbj profile image

drbj 6 years ago from south Florida

Fascinating hub, Aya. Thank you for the methodical research. It was of particular interest to me because before my first trip to Taipei, I attempted to learn enough Mandarin to communicate my basic desires for food, shelter, inexpensive souvlenirs, etc.

Gave it up though when I discovered that many Chinese words have up to five different different ways of pronunciation. Settled for learning just the Chinese phrases for "Where is my hotel?" "Where is the restroom?" and "That's too much money."

Aya Katz profile image

Aya Katz 6 years ago from The Ozarks Author

CaribeM, thanks so much for your comment. This article is written from my perspective as a linguist. But as far as speaking Chinese, I still have a long way to go. I did spend three years living and teaching in Taiwan. If you are interested in the cultural perspectives, in reference to the language and in general, here are a couple of hubs that you might like:


Aya Katz profile image

Aya Katz 6 years ago from The Ozarks Author

Drbj, thanks! Mandarin is not an easy language to learn, if you have never spoken a tone language before. Did you enjoy your time in Taipei?

tim-tim profile image

tim-tim 6 years ago from Normal, Illinois

Wow! I am speechless! You are so talented! I don't speak Mandarin but understand a little bit. It is not like Cantonese at all. That is why I still cannot speak Mandarin. Kudos to you!

Aya Katz profile image

Aya Katz 6 years ago from The Ozarks Author

Tim-Tim, thanks! But my Mandarin is not so good. As a speaker of Cantonese, you have a big advantage over me. Your language has more tones than Mandarin! So Mandarin tone should be easy for you.

June Sun 6 years ago

I would say that "friend" used as a verb is a sudden change. As a non-native speaker of English, I found myself using the compound "facebook" as a verb too: "Hey, can I facebook you?" It immediately made sense to the people I threw the question at. I would say that because there are many examples of a "word" used as a verb and a noun (zero conversion) at the same time, thus analogy is easy to take place for the the other words to be used as a verb/noun suddenly. We can always be as creative as we long as the creation makes sense to the speakers. :)

Aya Katz profile image

Aya Katz 6 years ago from The Ozarks Author

June, I agree! I think the use of a noun as a verb in English can be a sudden and instantaneous change, which will be adopted by others if it seems natural to them. The question then is: are English words underspecified for category?

CaribeM profile image

CaribeM 6 years ago

Aya Kats, thanks for the links to your other Hubs, both very interesting. I think you are being too modest about "not" speaking Chinese. ; )

Aya Katz profile image

Aya Katz 6 years ago from The Ozarks Author

CaribeM, thanks! You flatter me. Although I would not claim to be actually able to speak Chinese, I have never lost interest in learning. Perhaps it's never too late to improve!

    Sign in or sign up and post using a HubPages Network account.

    0 of 8192 characters used
    Post Comment

    No HTML is allowed in comments, but URLs will be hyperlinked. Comments are not for promoting your articles or other sites.

    Click to Rate This Article