Classical Chinese is taught in American colleges, if at all, with the greatest reluctance. Students are required to first take two or three years of modern Chinese, and only then exposed to classical texts. These are elucidated almost exclusively by bad translations into the modern language --- as though classical Chinese were nothing but modern Chinese with a lot of words left out.

Not only does this method not teach classical Chinese, it rarely works even to impart the modern tongue. Because of the difficulty of memorizing the characters, and the added trouble of all the words being monosyllables that sound like nothing in particular (though often all too much like each other), learning Chinese is at best tortuous. One must review and re-review the same material, till it seeps at last, drop by hard-won drop, into long term memory. Now, if the material one is passing ever and again before one's eyes consists of banalities that, once mastered, will only enable one to annoy people in restaurants --- well, most men of imagination simply quit. Even if one does persist for years, and chokes down all the ugly simplified characters and the diet of trivial chatter, one is still quite unable to read a Tang poem or a line from Confucius.

How much better it would be to begin with the classical language! Because of the great simplicity of Chinese grammar, one could plunge into real literature on day one. There is no long apprenticeship (as there is for Indo-European languages) during which one must memorize noun and verb paradigms. One can learn the grammar with the vocabulary as one goes along. But instead students are made to suffer through years of conning how to buy bus tickets, ask directions and tell time (to list the principal marvels imparted.) Only then may they be exposed to a real literature they which could have read at once. This is a piece of malignant stupidity which one would be hard put to parallel.

There is more. Not only is the student deprived of the immediate and easy rewards of his work, the hard work itself is made far harder. Memorize one must, there's no way around it. But one can get the same result by memorizing poetry and aphorisms rather than random vocabulary lists. Which is easier, which is likelier to live in memory and enrich the learner's life? Likewise the writing system is acquired with far less trouble via the traditional characters. Admittedly these have more lines, but they carry with them the traditional visual etymologies. More often than not these give accurate glimpses into the concepts the characters express. In the cases where if they are not true histories of the concepts expressed, they are at least a rich source of folklore. Even if these two great intellectual treasures were lacking, the traditional etymologies make the meanings of the characters, and often the sounds as well, instantly memorable.

Even if one's ultimate goal is knowledge of the modern language, this can be more easily attained via the ancient tongue than by starting with modern conversation, for the simple lexical reasons noted above.

I shall here present a concise grammar of the language, followed by a chrestomathy of classical texts chosen for their intrinsic interest and literary merit, with the object of giving the student material worth committing to memory. I assume a knowledge of pinyin and pronunciation, and the ability to copy characters, such as one may acquire from the first chapter of any good modern Chinese textbook.

Characters should be looked up at the Zhongwen website for the traditional etymology, and then looked up in The Chinese Character Dictionary for the radical. (See the Dictionary section of my Bibliography page for links to these resources.) The radical is what makes it possible to look up the character in the radical list of Mathews' dictionary (The alphabetizing of entries in Mathews is not easily used.) Mathews, whatever its shortcomings, remains the authority for meaning, since there is at present no better Chinese-English dictionary for the classical language. Never trust exclusively to the definitions Internet dictionaries supply.

The texts in the chrestomathy will be given in traditional characters, pinyin, literal and literary translation. A grammatical commentary will accompany each. I will attempt to reply to queries directed to me here.

This is a work in progress, which I shall be updating on a regular basis. The chrestomathy is as yet unposted. The grammar I post now, even in its incomplete and provisional form, since the clarifications on the use of yě and the construction of complex subjects is vitally important to the reader of classical Chinese, and nowhere else to be found. I shall be grateful to anyone who brings to my attention errors omissions. I may be contacted here.

This version was posted on March 27, 2008.

The Chinese themselves did not make a scientific study of their classical grammar, such as the Greeks or the Indians did. The traditional Chinese approach to the language is purely lexical. One learns the special idiomatic uses of the words by studying examples. Because of the great simplicity of Chinese, this approach is effective for handling most, though by no means all, points of syntax.

Traditional instruction in the Chinese classics is pretty much the same as what you'd find if you studied the Talmud at a yeshiva. The teacher reads and paraphrases into modern idiom as he goes, and sooner or later the student picks it up. Eventually the student develops "a feel" for how the classical language works. The merit of this approach is that it favors intuitive understanding. The shortcoming is that real precise understanding will often remain elusive.

The scientific study of Chinese begins only with the grammars written by Christian missionaries in the 17th century. Though our knowledge of the modern language has made great strides, and is now close to being complete, the grammar of classical Chinese has been neglected. This is due both to the Chinese' own indifference to it, and the purely mercantile motives of nearly all modern western language teaching.

Edwin G. Pulleyblank's Outline of Classical Chinese Grammar is the standard, and indeed the only serious work on this subject. I have relied upon it heavily for my own work, but even this fine book has a number of defects which it will be instructive to point out.

First there is the sheer obscurantism. Really, there is no reason to use words like "anaphoric" and "transformationally" if the intention is to be understood. And saying "adnominal" instead of "adjective" is just silly. Another vice from which Pulleyblank is not entirely free is "linguistics." The linguistic approach is to describe a language with meaningless elaboration, rather than prescribing and establishing the clear rules which a language will necessarily obey. Finally, Pulleyblank continues the lexical tradition of the Chinese, and most of his book consists of overblown dictionary entries, rendered less useful by the above-mentioned obscurantism and linguistics.

Pulleyblank's is the most useful grammar book I have found for Classical Chinese, and I am sorry to take it so seriously to task. But since it suffers from the above defects to the extent that I now have to write my own grammar of Classical Chinese, I think I may be forgiven a little plain speaking.

All Chinese texts are given in Pinyin, Chinese, and literal translation. The object of these translations is to make it transparently clear to the student what the Chinese means, thus I will always prefer awkward English to a smooth translation that hides the "watch-works" of the sentence. For the translations that appear with each quotation in Chinese:

I shall endeavor to always indicate words that I supply by enclosing them between square brackets [ ] .

Sometimes I have had to paraphrase, because the actual word-for-word this would have been, by itself, unintelligible. In such cases I shall use parentheses and the word "literally" ( literally: " . . . " ) to show what the Chinese really says.

In translation portions I shall give Chinese names in my own style of transcription, which is simply the best approximation of the sound which decent English orthography allows. Pinyin should only be used in transcriptions: to foist its mysteries on the unprepared reader is just scholarly malice.

Though Chinese is not an easy language, it is a very simple one. Once we grasp the full extent of this simplicity, and the problems which this very simplicity causes, we arrive at a truly scientific understanding of the language, and our work in learning it becomes immeasurably easier.

Chinese does not mark person, number, tense or case in the forms of its nouns and verbs. This simplicity in the form of the language entails that word order is rigidly fixed. The following rules are the very essence of Classical Chinese syntax:

a) this order may be inverted in an exclamation

3) A verb must precede its object, but

a) an object or other post-verbal element may be fronted for emphasis; b) interrogative pronoun objects precede the verb

c) pronoun objects are placed just before the verb when the verb is negated

4) Normally the subject precedes the verb and the object follows, but

Easily inferred words are frequently omitted, and so the most fundamental part of learning Classical Chinese is coming to know what to supply. When a word which is typically omitted is in fact given, we may assume that it is so to give additional emphasis, or to avoid ambiguity

Subjects are more often than not unexpressed in declarative sentences. This is particularly the case when:

the subject is easily understood from context, ("He ran home, he picked up the newspaper, he scanned the ads.")

when it is indefinite ("One should always do one's best"),

and when it is impersonal ("It's raining.")

All tenses are expressed by the present tense form of the verb. Other tenses must simply be inferred when required.

Often, we must not only give the verb its subjects or objects, but also supply the conjunctions. Typically omitted are the conjunctions which give one verb an adverbial relation to another: when, because, although and if (and of course their synonyms.) Coordinating conjunctions, like and and then must also, ordinarily, be supplied.

There is no "that" in classical Chinese to mark a subordinate clause. Thus the second Chinese verb in a sequence may begin a subordinate clause. For example, it may indicate the purpose of the first verb, in which case we must preface it with a phrase like "in order to." The subordinate clauses are actually learned rather easily, by the traditional method of studying the Classical Chinese with a modern gloss. Recognition of these structures is easy for us because English already expresses subordinate clauses of purpose, result, intention and so on without any special tenses or introductory particles, usually just by adding an infinitive ("He worked hard to get into medical school." The idea here is a purpose clause: "He worked hard in order to get into medical school.")

The particle yě may always be omitted where it is clearly implied. This is true with nouns as with verbs. The use of yě is discussed in detail below.

The only interesting problems Classical Chinese poses to the western linguist arise from the ways in which it truly differs from an Indo-European language. These appears in the ways it supplies its most pressing lacks: relative pronouns, and infinitive and participle forms for making nouns out of verbs. For these matters, example is not better than precept and we must carefully analyze the structures.

There is no distinction between singular and plural for the pronouns. All pronouns may be used as subject and possessive. Those that can also be objects, are emphatic when used as subjects. These latter are indicated.

Archaic and rare even in classical writing:

Common in classical writing:

The possessive meaning is particularly common, and differs from that of qí. Zhī is a reflexive possessive: it means, for example, his/ her/ its/ their own. Qí is simply a possessive, and may mean, for example, his/ her/ its/ their in the sense of someone else's. English does not have a reflexive as opposed to a possessive pronouns (as Latin does with suus as opposed to eius). Thus in English we often have to add parenthetical information to clarify our meaning. We might say for example, "He lent his (his own) car to his (his friend's) brother for the afternoon. Chinese has no such ambiguity:

This is a particularly interesting example. Usually the zhī comes immediately after its subject, and amounts to an apostrophe-with-"s" ('s). The quote above makes it clear that zhī is a reflexive possessive, and refers back to its subject from whatever position it occupies.

As object:

When the language or thought of any person is reproduced without change, this is called Direct Discourse. Direct Discourse is the quote that goes inside quotation marks. When, on the other hand, when someone's language or thought is made to depend on a verb of saying, thinking, or the like, this is called Indirect Discourse.

For example:

Direct Discourse: The master said, "This is what is expected of a gentleman."

Indirect Discourse: The master said that this was what was expected of a gentleman.

In Chinese the range of verbs that can introduce Indirect Discourse is rather wide.

Any verb may be used as a causative, and the causative use of the verb extends to indicate how something is considered.

Yě functions as the verb "to be." Though frequently omitted, it is usally expressed to say that something is something else.

These are the primary uses of yě, on which all others depend.

Yě is used whenever we want to say that something is something else. For example, "Socrates is a man" would be expressed:

Yě may be placed after a noun to introduce it as the subject.

Literally, Lǐ yě means "[It] being Li." As usual in Classical Chinese, the subject ("it") is not expressed when it is impersonal. (see Basic Principles, Omissions, above.)

A single yě at the end of a sentence may serve as the verb "to be" for several preceding phrases as well. In this way it helps mark the end of the clause.

Classical Chinese also omits the yě with numbers, which in English require the predicate nominative form. (For example, "The profits were fifty dollars.") Bear in mind that numbers are adjectives!

Yě is omitted where it is clearly implied. Certain words create the anticipation of the verb "to be" at the end of the sentence. In such cases, the ye3 may be omitted. This is frequently the case with exclamations and questions. The merest hint of special emphasis suffices to suggest the verb "to be."

Thus yě is often omitted with

Zhī qí zuì zhě wéi Kǒng Jù Xīn.

The only knower of his faults is Kong Ju Sheen. (Mèng 1B/4)

In a conditional sentence this is also often the case.

This is also true of the imperative, which enjoins something for the future (albeit sometimes the very near future.) Neither Ye3 nor yi3 are ever used with imperatives.

Frequently a single use of yě does double duty as the verb "to be" and as the indicator of ongoing action. In such a case it must be translated "being."

The yě modifies only rén. Literally, rén yě means "being a men," and is equivalent to "who, being private persons". The yǐ at the end of the sentence applies to the wěi alone. (Both yǐ and wěi are discussed in more detail below.)

Yě after several verbs modifies them all and indicates the conclusion of the sentence. However:

Yě is to be taken with the modal verb where there is one, and not with the verb that follows the modal Thus yě serves to bracket the entire modal verb phrase, and helps define the units of meaning in a sentence.

In the first half of this sentence, the ye3 goes with the neng2, and emphasizes that the impenetrability of the shield is an ongoing condition. It does not serve as the verb "to be!" Non-existence of a thing that could pierce the shield is taken care of by the mò, "[there is] nothing." Mò cannot take a yě. In the second half of the sentence, a néng is implied between the bú and the xiàn. (Note that wú, " there is not" cannot take a yě.) As always in classical Chinese, anything that can be inferred may be omitted. But the inference must be clear, as here, where the perfectly parallel first half of the sentence gives the pattern for the second half.

A possessive makes a separate phrase which is end-bracketed by yě.

A gerund is a noun formed from a verb. It may still act as a verb (for example, by taking a direct object). In English, a gerund ends in "-ing" and the -ing ending provides the best way of translating this use of yě.

No second ye3 is added when a final ye3 forms a gerund which is also a predicate.

The final yě both makes the gerund "not-being-able" and supplies the "is." Note that use of fēi shows that "not-being-able" is a verbal noun.

The gerund may be the direct object of a verb.

Here the gerund yǐ . . . yě takes the direct object zhī.

But the presence of the possessive before the gerund may make it it clear enough that a gerund is required and so the yě may not actually be expressed. The logic here is this: while a subject introduces a verb (for example, "he thinks"), a possessive introduces a thing (his thinking). Since the gerund form is required by the possessive, it may be taken as implied and need not be expressed. The following example is instructive:

The idea is that the imperial ritual is so profound and rich in meaning that one who understood it would also understand all else, and be able to reveal it as easily as one shows one's palm.

Now the first yě, which one would expect after zhī shuō is unnecessary because the zhī implies it. Next we have a yě which is given because Confucius wants to stress that his ignorance of the meaning is an ongoing condition. He does not know it, and he does not expect to. The last yě, which is given, makes a gerund ("being related") out of the preposition yú; it was not strictly speaking necessary, since the possessive zhī implied it, but it was added for clarity since making a verbal noun out of a preposition is a less usual proceeding.

Note that Chinese prepositions behave like verbs in many ways, which is why Sinologist call them (horribly enough) "co-verbs."

In the Analects we find

Wéi, once again like the Spanish estar, may indicate a condition which is only temporary in contrast to a truly permanent state:

The contrast here is between what one is (which is permanent) and what one does (which is temporary.)

In a logical extension of its completion meaning, yǐ may also give an emphatic sense to a verb. The notion is, the action is not only complete but completely done.

At times yǐ is used purely to intensify some word which has no real verbal meaning in play:

If the yǐ had gone in its normal place, at the end of the sentence, we would have translated it with one "very." Putting the yǐ after the xiān puts extra emphasis on that word and thereby makes a final yě or yǐ unnecessary. (See omission of yě above).

Even though this structure is often used to express something one has just become aware of ("Look at the clock, how late it has become!") it does not necessarily indicate a change in awareness only, as Pulleyblank would have it.

The particle zhě makes a concrete noun out of an adjective or a verb. If we absolutely had to translate it all by itself, I could probably not do much better than to render it " -entity."

Here two adjectives become nouns to describe an inkeeper's two wives.

Of course in English we would use the word "one" in place of entity in a case like this: "the ugly one . . . the beautiful one."

Zhě may also be used to create what amounts to a relative clause:

In English we would express this with a relative pronoun: "a friend who is not like yourself."

Another possibility is to use zhě,

It's easy enough to use these structures to say something like "the man who came to dinner." But if we want make a relative clause as long and complex as "the man whom I mentioned to you earlier in such glowing terms," we have to use more elaborate means

In Chapter VII, Noun phrases and Nominalization, Pulleyblank makes a very complicated business out of the two very simple structures used for this purpose. He derives the phrases from supposed proto-phrases and represents it all in algebraic style. This is the very madness of linguistics.

Here's how matters actually stand. Take a phrase like

Very literally this is "a kill-man of a king." We should translate it "a man-killing king." (The Chinese formulation may make more sense if we think of the parallel English idiom with "of", for example in the phrase "a hell of a good cheese" which we would put into proper English with "a hellishly good cheese.")

Here is how the Chinese phrase works in detail. The verb-phrase "kill-man" is reduced to the status of a mere modifier of the word "king" by the possessive zhī. So the phrase means "a kill-man king." The part which is puzzling to English speakers is the zhī, which we are used to translating as "of." But here the meaning of the zhī is not possession. Rather, it shows what the phrase modifies. We could translate the phrase concept-by-concept thus: "a kill-man/ this-phrase-modifies:/ a king."

Now the English equivalent of a verb phrase which modifies a following noun is a participial phrase, like "man-killing." We should leave the "of" out of our translation because it is only there to indicate the subordination of the verb phrase to the following noun. So we translate the Chinese phrase sha1 ren2 zhi1 wang2 thus: "a man-killing king,"

Besides the above construction which I compare to a participial phrase in English, there's another fancy way of making a subject in Chinese:

Very literally, this reads "a king of a kill-men entity." We should translate it "a royal killer." As with the verb phrase "kill-men" above, the noun "king" is reduced to the status of a mere modifier by its placement before the possessive zhī. A noun which modifies a following noun acts as an adjective, and the adjective we make from "king" is "royal." As before, we leave the "of" out of our translation because it is only there to indicate subordination of modifier to the modified.

Now that we see that wáng zhī is correctly rendered by "royal. It remains to show how shā rén zhě yeilds the meaning "man-killer."

The second part of the the complete phrase, the modified, is the verb phrase shā rén, "kill-men." This is turned into a concrete noun by the zhě that follows it. I express this in ultra-literal translation by adding the word "entity" ("a kill-men entity.") But we can express this idea more idiomatically. Zhě changes an action into an agent, just like the suffix "-er" in English. Thus it changes the phrase "kill-man" to "man-killer." Putting together the part that amounts to an adjective modifier (wáng zhī) and the part that acts as a noun modified by that adjective modifier (shā rén zhě), we arrive at the translation "a royal man-killer."

And so we see that Classical Chinese has two ways of making a complex subject. We can use just zhī between the verb-phrase and the noun to say "a man-killing king." Or, by putting the noun ("king") in the subordinate position, in front of the zhī, and adding zhě, we produce the phrase "a royal man-killer." The two formulations are interchangeable. The only difference is that one ("a man-killing king") emphasizes the fact that he's a king, and the other ("a royal killer") that he's a killer.

Thus Chinese, using zhī, can make longish phrases into complex subjects. The examples about the homicidal king given above are unusually simple. The following instances show the elaborated use these structures actually enjoy.

Here the phrase contains a nice pile-up of information, including a reflexive verb.

The two verb phrases "do-worthy" and "set aside" are subordinated to "heart" by the zhī. Note that qū, "leave," is used in a causitive way, and that the reflexive pronoun zì has the peculiarity of preceding the verb that takes it as an object.

This modest list is comprised of items for which there is a classical usage which is not completely described in Mathews. In cases where Mathews gives the best and fullest description, the citation in Mathew is given after my brief note of the most common meaning. Where Mathew's statement of the meaning is brief and helpful, I give this in full, immediately, in quotation marks.

As a prohibitive, it means "shouldn't." "let . . . not." (Apparently the "have" meaning implicit in wú supported the addition of an optative sense.) The conceptually simpler wù (below) acts as a direct command in the indicative (fact, not possibility) mood. Accordingly, wú is frequently used in conditional sentences ( "If he does not . . . then he will . . . .")

Zhū is used as an adverb in this famous line, where the shift in meaning from "all" to "entirely" is easy to understand.