At the beginning of this century, Chinese reformers borrowed the ancient substance-function distinction for a popular slogan: 中 學 為 體 西 學 為 用 Chinese learning as the substance, Western learning as the function." This was a formula for blending Chinese and Western culture. Chinese learning is a political-moral dao and Western learning is science and technology. China, according to this slogan, should preserve its substance of moral and political insights while using Western scientific and rational techniques.
Where does the rule of law fit into this equation? One obvious important change at the beginning of this century of interaction is a near consensus in "greater China" that Chinese political culture should adopt the rule of law.  Even the top salesman in the region for Asian-values, Lee Kwan Yiu accepts its importance.
But this consensus masks a puzzle. Orthodox histories of China tell a story according to which China tried the rule of law in the Qin dynasty. It was developed and advocated by a school known as Fajia and a standard translation of this term is "Legalist".
To be sure, those histories also note that after the brief "experiment" the political culture quickly repudiated Fajia ideology and instead chose Confucianism as its political-ethical "substance." However, those histories also note that the actual practices put in place in the Qin dynasty, the legalist political institutions, survived in Chinese dynastic rule right up to modern times. So we have the following orthodoxies:
1. Chinese Legalists 法家 founded a political system based on law.
2. The culture repudiated Legalist philosophy but continued its political structure.
3. Conclusion: China already has the rule of law. It should have no need to adopt it.
One of these assumptions must be wrong for the consensus to make any sense. What is there to adopt? If we somehow explain what happened, we still have a further question. Why should Chinese culture make a different decision today from the one they made after the Qin? If we (speaking as a Hong Kong resident) adopt rule of law now, how do we deal with the problems that lead to abandoning it originally?
I posed this puzzle years ago in an earlier discussion of law in China.  My conclusion then was that the orthodoxies were mistaken. China never had a normative theory justifying the rule of law. I thought then, as many do now, that the Legalists in China had argued for a rule by law but not rule of law.  China's theoretical rejection and continuing residual practice concerns only the former. Chinese thinkers never had a normative insight into the value of rule of law.
This distinction by preposition is familiar in discussions of Chinese attitudes about law. I characterize it as an attitude toward the role of law, its purpose, its justification and norms of its application and interpretation. 
Rule by law
Rule of law
Control people and achieve social order by promulgating general commands backed up by punishments
Control officials and rulers by restricting their use of the state monopoly on force
Law as the dictate of rulers controlling the ruled 
Give people clear guidelines about what actions are liable to punishment
These are not different kinds of legal systems, but different theoretical accounts of the normative role of lawdifferent norms of commitment to the institution. What I assumed then (with most interpretive theorists) was that Chinese legalist thinkers had argued for using punishment as a way to control the people. The basis for control was the arbitrary commands of a ruler rather than either morality or Confucian ritual. the Legalists, however, had neither argued for the primacy of law over the rulers or officials nor that government should be restricted or regulated by law.
I now want to defend the nominal orthodoxy, but with what I hope will be a more interesting twist. What I want to argue here is that the philosophers of ancient China did produce an little-noticed but viable moral argument for the rule of law. The way that argument works is surprisingly akin to the modern, liberal or human-rights justification of the rule of law does. This argument should thus undermine the line that suggests that traditionally Chinese conceptual structures or fundamental norms were such as to make impossible their appreciation of liberal arguments for the rule of law.
Still, I will argue, the Legalist implementation of the rule of law conceived in these terms backfired. The contemporary Chinese analysis of this failure that argued for abandoning legal institutions also presupposed this norm of what law was supposed to achieve. When we see why they perceived the experiment with the instution as having failed, we will begin to see the importance of a Western technology to achieve the shared goal envisioned for law in both traditions. The technology is the Western interpretive technique that mimics scientific theorizing. The issue of law, in other words, neatly fits the mold of Chinese moral substance being enhanced using Western scientific function.
Let us begin with something I was right about in my earlier analysis. Strictly speaking, Chinese thought in classical time did not have a concept of 'a law.' In political contexts, we easily to forget that the concept of "a law" is not the exclusive territory of lawyers, judges and government. The Western concept of law is not equivalent to "penal code." Penal codes are one relatively narrow instance where we employ the concept of laws. Scientists, logicians, and moralists have equal (if not prior  ) claim on the use of this Western concept. We speak of laws of nature, laws of thought, moral laws as well as "legal" laws.
A law, in this broad sense, is something we characterize in broadly syntactic and logical terms. A law is something
(a) we can express in a sentence
(b) with a universal element (e.g., 'all', 'no' or lawlike 'if then ')
(c) and some type of necessity
Mere accidental universality does not make something a law. To count, a universal sentence must not merly be true, but have a strong modality of the type appropriate to the system of reasoning. Scientific laws have descriptive or causal necessity. Moral laws have prescriptive necessity (duty).
In all of these areas of use, we view laws as crucial elements in a scheme of justification, explanation or proof. The logical or inferential nature of the practice corellates with the different senses of necessity that its laws have.
I still conclude that Chinese thinkers did not have this broad, syntactic, logical and modal conception of law. Where I went wrong was in thinking that the concept of a penal code could not have served the normative purpose of motivating a conception of a well-ordered society as one governed by rule of (not by) law. The rule of law would play a role as a technique backed by a moral argument for restricting or limiting government coercion. I did not then see how an alternative, non-sentential, notion might have served in this normative role.
What this syntactical observation about 'law' really does best is help explain something about the development of the Western conception of law and the Western conception of interpretation. The formal notion of 'a law' invited Western thinkers, drawing on their Platonic tradition, to look for a common essence linking all the realms of reasoning that employed the term. As was typical in traditional Western thought, this motivated a search for the "definition of 'law'."  A famous definition of that shared essence among all systems employing the word was that of an ordinance of reason. 
This definition worked alongside the dominant religious theories and the emerging "natural philosophy" to tie the three realms of laws togethera knot that lasted until the "early modern" period. Medieval, religiously oriented thinkers portrayed law as a continuous, coherent and unified body of divine principles: the rules of the supernaturally rational creator--the mind of God. Western philosophers, as a result, had a religious-metaphysical basis for distinguishing human penal codes or commands into two categories. Some would be valid laws and others invalid (borrowing the shared language of logic). Valid penal commands are those consistent with divine command and purpose (i.e., morality). A valid law was a dictate of reason for the well-being of God's favored creatures--us.
Natural law's unity began to be broken when science split-off from religion. Some legal theorists (positivists) rejected the religious view that validity of human law was dependent on its derivation from God's eternal law. They retained, however, the broad "dictate of reason" conception of law. John Austin characterized law as "a rule laid down for the guidance of an intelligent being by an intelligent being having power over him."  The positivists tried to purge the concept of law of any normative content. Toward this intellectual goal, they slipped into a conception of the role of law that coincides with rule by law. The implied role of law was to control the people using dictates or commands of the ruler. It is this tendency that invites the charge that positivists give no basis for fidelity to law.
Natural v. Positive Law
All law as dictates of supernatural reason
Laws as arbitrary dictates of rulers'>
Causal, moral & valid penal law are coherent'>
Separation of law and morality
Knowable by and for (rational) humans'>
Valid law means only that one is likely to be punished
Much of the discussion of Chinese law  turns on two related issueswhether Chinese thinkers treated law as divine in origin and whether they had a concept of natural law. I believe that both of these are relevant, but relatively tangentialsymptoms, not the underlying condition. Divine origin is not the crucial belief in explaining this important feature of Western thought. The key is the conceptual link between law and reason. This link provided the conceptual glue that held together different realms of "necessity" (scientific and normative) and pushed natural law theory in the direction compatible with Western religion. It resulted in an account of law, according to which, rational humans (in the image of God) are capable of distinguishing valid from invalid law.
Similarly, though I agree that the natural law tradition is important in explaining this development, the fact that the law has a metaphysical status (rooted in nature rather than convention) is not crucial. The key to understanding the normative turn in Western legal theory is that the faculty of reason apprehends the normative status. That same faculty guides "natural philosophy"science and scientific understanding. Reason's pure subject matter is logic (consistency and coherence). That subject matter essentially concerned the inference relationships among sentencials. While disagreements about what normative or descriptive law is abound, disagreement about logic should be inconceivable except as gross error. Logic is laws of thought.
In China, the accessible, theoretical candidate faculties for grasping "natural law" would be the moral intuition. No concept of logic, of norms of inference among sententials, was available, so what has to be grasped is potentially controversial norms. Absent the unifying concept of 'a law' and the peculiar religious story involving reason, Chinese thinkers had little ground to regard penal codes themselves as having any grand legitimacy. They came to view the choice, as Western positivists did, as between different norm systems. Morality  and penal codes formed an either-or choice among independent and distinct dao. This led the Chinese perspective initially in the direction of the Western legal positivists who reject natural law and insists on a separation of law and morals and and who analyzes legal duty as little more that a liability to punishment. Let us look more closely, however, at the Chinese formulation of that choice and the analysis and response of ancient Chinese normative theorists. A powerful Confucian argument for rejecting rule by law required those defending penal publication to think of its role differently. I will argue that this drew on concepts used in the Mozi and resulted in an argument for rule of law.
Chinese thinkers certainly were cognizant of legal codes, whether or not they had the formal sentential concept of 'a law.' A great deal of debate in ancient China concerned whether or not penal codes should be publicized.  The interesting point, as Bodde  notices, is that these discussions suggest Chinese were thinking only of penal law. Again, this is significant, but, I will argue, a symptom of something more fundamental. That law concerns criminal punishment cannot, by itself, block a moral argument for the rule of law. The importance of Bodde's insight lies rather how the genesis of the debate shaped our different views of the conceptual structure and role of penal codes.
The dominant Chinese term for these codes in the early debates was the notorious xingpunishment.  Our identifying them with legal codes tempts us to picture them as having the form of commands from the ruler to the populace. The publication that so offended Confucian sensibilities, however, was apparently conceptualized as a specification of (appropriate) punishments.  The identification of offences was a by product of codifying the punishments, not the primary purpose of the code. In other words, they did not treat the code as having created legal duties or offenses. As they were initially conceptualized in philosophical argument, the codes were not taken to prescribe behavior directly. They directly validated and approved a range of punishments for offences. The debate did not present the codes on the model of a set of universal sentences creating legal duties. It guided or regularized punishment.
This feature of the debate should help us better understand the Confucian reaction to publicizing the codes. Consider this famous argument from the Analects:
Lead them with政 zhengcoercion, order them with刑xingpunishment, the people will avoid [wrongdoing?] but will have no shame. Lead them with德 devirtuosity and order them with禮 liritual and they will have shame and learn to fit in. (2:3)
I particularly like this argument from the Analects because it is a rare example of a pro-authority philosopher arguing against rule by law.  Western philosophers seldom questioned that justifying political structure was justifying law. Typically, only anarchists opposed law (as an adjunct to opposing government). Confucius, however, plainly was no anarchist.
We can easily read Confucius' argument as a stinging indictment of Western, law-ridden societies if they posture as models of the advantages of positivist rule by law, i.e., if they take the justification of legal structure to be that of maintaining social order. Reliance on penal codes and punishment exercises and strengthens the selfish tendency in human nature. Legal codes stimulate people to think in terms of preventing harm to themselves.
If one succeeds in talking his way out of the punishment, he will be an object of admiration and the next person who gets in trouble may seek him out. For a pair of pants, he helps that person before the magistrate and maybe can get him off too. The ruler has to make the code more complicated, then the sophists get cleverer, and we soon all lose our shirts as well. Eventually we have a society filled with self-interested people who fund a huge army of lawyers and lawmakers and law libraries trying to keep up with an ever more complicated legal code. Even the political rulers will say "What I did was immoral, but not illegal."
Ultimately, a society that relies on such an institution for its social order will not be as safe and orderly as a society that relies on moral education and building people's character. It is better in the end to depend on and exercise the equally natural inclination to social conformity and traditional norms. Then you will have a citizenry that "knows shame."
Relying on penal codes as guidance for people plainly conflicts with the Confucian emphasis on instilling moral character via education. As the Mencius effectively argues, "When they commit crimes, to catch and punish them is to trap them. How can something like trapping people be consistent with the rule of benevolent man?" (Mencius 3A:3) To publish codes and punish people without teaching them how to behave and cultivating their character, is a cruel game. Penal codes without moral education provides no social guidance. We contrast Washington D.C. and New York with Hong Kong and Tokyo. We are not safer here on the streets of Hong Kong because we have more elaborate penal codes, but because we have a Confucian system of family education and character cultivation.
The second edifying plus to this Confucian argument lies in how effectively it undermines rule by law. The argument for rule by law assumes the justifying role of legal institutions (published penal codes and mechanisms of punishment) is to guide human behavior. Confucian opposition rests on the charge that law is neither necessary nor sufficient for such a role. It forces anyone advocating those institutional arrangements to produce an alternative account of their role and purpose. Confucians share the consequential goal of social order, but Confucius here makes a strong prima facie case that we get a stable social order more reliably in other ways. Advocates of rule by law are unlikely to be able to meet the argumentative burden of showing that punitive institutions yield social order more effectively than any other dao. Given the obvious consequential evil of cruel punishment, failure to meet that burden is fatal.
Of course a practical politician, as opposed to a political philosopher, can plausibly argue for the short-term advantage of penal codes over education.  As we will see, however, in China the line of argument turned more on (a) the flexibility and adaptability of promulgated public guiding codes in changing conditions and (b) on how penal codes could advance programs of social reform. They rarely relied on the simple contention that socialization took too long to bring order to society. The appearance of appeal to short run considerations masks a reformist justification of public codes which originates in a very un-Confucian view that established patterns of social behavior should be changed.
The story of Confucius' case against rule by law as I embellished it above, has two components worth distinguishing. One is psychological and the other interpretive.  Confucius' argument in the cited passage technically invokes only the psychological half of the narrative. The psychological argument is that punishment stimulates selfishness and impedes the natural growth of a socially harmonious character. The interpretive argument, by contrast, is the part that funds our shared hostility to lawyers. It is a more complicated and problematic attack on the public accessibility feature of the codesthe specific objection to their publicity as opposed to the use of punishment as a technique in ruling.
For those anxious to preserve respect for Confucius, I note that I cannot find an unambiguous commitment to the more controversial interpretative objection in The Analects itself. His quoted statement presupposes only the psychological theory. Hints of the interpretive objection surface elsewhere in Confuciuss Analects, e.g. in its occasional denunciation of "glibness" and its distaste for "litigiousness."
However, for explicit invocations of the interpretative argument, one has to go to other Confucian texts, e.g., this famous passage from the Zuozhuan.
A normatively embarrassing feature of this objection is the classic "rule-of-man" assumption that an elite group of authorities has cultivated, intuitive, normative insights which are more accurate and reliable than those of the people. The objection notes that reforms introduced via penal codes undercuts respect for this traditional moral elite, it undermines their authority. Such appeals could hardly be valid when reformers are specifically proposing equal treatment under the law. Similarly, the implicit complaint that reform leads to abandoning traditional ways is question begging in this context surely that is precisely what the reformers intended.
A related worry plagues the interpretive argument. It may also beg the question vis a vis all the rival schools in China. Rivals would vehemently deny that Confucian gentlemen are morally superior or that their 禮 liritual is the legitimate standard of correct behavior. The skepticism about claims to such moral authority marks the dawn of philosophy in China. Each of the "hundred schools" has a different program to reform tradition so each implicitly rejects the Confucian gentleman's posture as a moral authority. Conservatives could hardly expect their rivals to accept the relevance of their complaint that such reforms undercut the authority of Confucian gentlemen.
However, we can extract a more respectable general worry from this argument. It lies in a striking grasp of the philosophical difficulties of (legal) interpretation. "Lawyers" can find loopholes, conflicts and exceptions in any code. This worry fuels Confucians' deep distrust of "name-splitters"  and becomes a recurrent theme in ancient Chinese thought. It arguably undergirds the Analects complaints about "glibness" and "litigeousness." This makes its most interesting and important appearance in its famous account of 'rectifying names.'
This passage from the Analects, however, undermines the Confucian version of the interpretive objection to public penal codes because it soundly acknowledges that, whatever the interpretive problem is, it plagues conventional 禮 liritual and music just as it does codes of punishment. The interpretive problem is a general one in relation to codes that guide conduct. A code posted in public, one written in an ancient ceremonial text, and one in the form of musical notes all share this problem. We cannot apply any recipe for action unless we can pick out the correct referent of each term it uses. If we select the wrong ceremonial cap or the wrong ritual flask, then we simply fail to follow the ceremony correctly. If we misidentify "forgery" then we punish the wrong people. If I do not distinguish allegro from andante, or F from G, my performance is wrong. If we do not agree on what the terms in our code refer to, then the code cannot coordinate our behavior and attitudes.
Imagine the justices on the U. S. Supreme Court trying to decide whether to accept a law that punishes advocating communism. They work with the famous "basic law" provision "congress shall make no law . . . abridging the freedom of speech." They ask themselves if we should call advocating communism "free speech." Earlier courts have already ruled that obscenity, libel, and blaspheme are not "free speech." The Analects would treat this interpretive judicial activity as an example of "rectifying names." Its analysis of the interpretation problem makes is seem to defy solution. How do we know when we have rectified a name correctly? What makes the difference between the Confucian rectifying and the lawyer splitting or confusing names?
The Confucians do propose a solution--of sorts. However, while the interpretive problems faces both the ritual and penal code, their solution works for only one of them--the 禮liritual. Confucianism began to emphasize a concept of a moral intuition, 仁renhumanity, which humans share with the ancient Sages who wrote the code of 禮liritual. We imagine that those who have cultivated this intuition can just "see" what performance the sage kings would have intended in this situation. Confucius, for example, sees that the use of a cheap silk cap does not make the ceremony wrong, but eliminating a bowing step does. However, if an intuition allows us to mimic the state of mind of the sage kings, then, as Song-Ming period Confucian theorists realized,  it also makes the code unnecessary. We can operate (as the sages themselves did) with the intuition alone.
Thus, his intuitive sense of "how it should go" helps the Confucian sage "rectify names." This version of a moral intuition as something natural, shared by all humans, linked to the moral sense of the original founders of society and cultivated by education in a traditional morality, makes 仁renhumanity a dud in dealing with the problem of legal interpretation. Naturally, therefore, it pulls Confucian theory in the direction of the rule of man. It proposes an intuitive moral sense as the authority substitute for a public code of retribution. It implicitly values intuitive flexibility as opposed to fixed rules and it relies on the cultivated intuition of an elite scholarly class. 
The Mohists were the most systematic of those arguing for moral reform. If we want to reform our moral traditions, we need to have some standard that helps us decide what is "better" than our traditional moral judgments.  The standard must tell us both what to change and how to change it. The Mohists proposed utility as the standard--we change the traditions in ways that lead to the greatest overall social benefit.
In the modern West, social reformers also appealed to a utility test. They presented it in a mathematical form; Bentham called it a moral calculus. The Mohists were craftsmen--essentially carpenters--and they portrayed the benefit/harm standard as analogous to a measurement. A good example is a carpenter's plumb bob. If you want to apply the term 'below' or 'vertical' in the blueprint (the guiding code), you dangle a weight on a chalked string and then snap the string. The "operation" tells you how to follow the code.
Mohists said that, in contrast to cultivated, scholarly intuition, the operational measurements for language should be available "to the eyes and ears of the people." The Mohists called these 法 fastandard and argued that the utility standard was a fa--an operational, measurement-like standard for answering moral questions. We can use the benefit/harm measurement distinction both for rectifying names and for deciding what changes to make in a traditional moral, ritual or legal code. So, the Mohist fa is their alternative to the Confucian intuition, an alternative that makes moral reform possible.
Another ancient text, the Guanzi, takes a similar line, but without attacking Confucian禮liritual. The Guanzi troubles the traditional analysis because it seems to mix Confucianism and legalism. It champions both 法 fastandard and 禮 liritual. However, the puzzle disappears when we remember that fa is a rival standard of interpretation, not a rival code. So, the Guanzi stands for an lucid hybrid position between the Confucian and the Mohist. It accepts the traditional code, but not the use of scholarly intuition as the standard of correct application or practice of the code. The Guanzi's examples of fa, like Mozi's, are mainly objects used in measurement operations: balance scales, bushels, squares, compasses and so forth. 
In effect, we see here two rival positions emerging on the question of interpretationneither of which involves the concepts of meaning or definition. They are techniques for fixing reference in the course of acting. One position advocates reliance on an intuitive 仁renhumanity to 正 名zheng-mingrectify names while taking a historical, social-cultural 道 daoguide as given. The other position proposes some operational or measurement 法 fastandard as a way to guide interpretation of some 道 daoguide. The 法 fastandard approach appeals to reformers because it can also be used as a way to evaluate proposals for changing traditional 道 daoguide.
The Fa-jia, as their name suggests, took up this reformist idea. As Tang Chun-I most clearly recognizes, we need make no change to the Mohist and Guanzi view of fa to see the point of Fa-jia theory. The essential feature of penal codes was their publicity, their availability to the eyes and ears of the people. Further, Fa-jia thought the codes should have a measurement-like clarity and simplicity, e.g., if you cut off one enemy head in battle, you advance one rank. The example stems from accounts of Shang Yang's administration in the Hanfeizi. Shang Yang was the practical statesman credited with first systematically instituting fa in a state. The idea was that penal codes should be as clear to people as measurements are. The Shang Yang makes this surprising moral argument for such public, promulgated codes. This is the argument I had earlier failed to appreciate. 
The basic justification of clear, objective, measurement-like codes here is that they protect people from official oppression. If people know what the code is, the officials do not dare punish them except in accordance with the easily interpretable code. If people can claim that their action was not a violation of any clear published rule, the officials cannot punish them.
This line of reasoning leads a series of otherwise quite puzzling claims in a book attributed to one of the most vilified and demonized figures in ancient Chinese history. In addition to the suggestion that people "happily govern themselves," this texts avers that fa pacifies people. Another passage says that fa will eliminate punishment.  The implicit reason is manifest in this passage. Punishments are attached only to clearly specified actions which people know in advance. This gives them the wherewithal to choose a course of action that avoids official coercion. This flips the above, self-serving Confucian argument against public codes on its head. Indeed, the law will cause people to lose their fear of authorities since they can appeal to the word and ensure their immunity from punishment--and that is precisely what Shang Yang took as the reason for clear, objective, published codes restricting punishment.
There are several reasons I failed earlier to notice this line of argument in the Fajia. The first is the orthodoxy we all learn from Chinese-English dictionaries. It says that fa changes meaning when used by Legalists. Then it simply means laws, not objective standards for interpretation. The orthodoxy detaches legalist reasoning about fa from the goal of objectivity and easy accessibility to people.  It converts fa into an explanatorily otiose synonym for xing, merely another character for the same concept--penal codes. If we then let translating it as "law" invite us to project that fa matches the Western idea of a law, i.e., of a universal, necessary sentence, then this line of argument will be lost.
The orthodoxy tempts us to read these claims as assertions that punishment is what pacifies people. That claim has to strike us as naïve in the context of the pervasive worry about interpretation that lies behind the prior use of 法 fastandard in philosophical argument. The Fajia passages about how 法 fastandard controls officials are warped into simple claims that officials, like everyone else, are subject to punishment. The frequent reference to how法 fastandard limits their options to punish and reward ordinary people disappear from the explanation of how 法 fastandard benefits people. The almost ubiquitous worries about the vagueness of language and guidance and the well-attested and unrefuted argument of Confucians about motivation are simply submerged in simplistic appeal to cruelty. It turns the Fajia argument from a plausible one for rule of law into a dubious and already well-refuted argument for rule by law.
A second reason I missed the argument was the extremely negative image of Shang Yang in traditional accounts of Chinese thought. We do not look here expecting liberal arguments made from a point of view sympathetic to ordinary people. A third reason is that Hanfeizi dominates our conception of legalism. And Hanfeizi manages to distort this crucial line of reasoning almost out of recognition.
Hanfeizi does implicitly register the above line of argument but invariably recasts it so it reads purely from the ruler's point of view. He emphasizes how 法 fastandard weakens officials relative to the ruler where Shang Yang noted how it benefits people. The threat to the ruler, Hanfeizi argued, always comes from his high officials, not the people, so he aimed all of his techniques of rule at controlling the bureaucracy. If there are clear, public measurement standards then officials cannot use punishment and reward to foster loyalty and build a power-base. This twist still echoes the theme of the original, unifying Confucian argument against the rule of law and Shang Yang's "protection of the people" analysis. However, it locates it in an overarching concern for the ruler rather than for the people.
We have now encountered the core argument from three different goal perspectives: first from the point of view of the officials (The Confucian objection) then the protection of people from officials (Shang Yang), and finally the interest of the ruler (Hanfeizi).
Hanfeizi was the most well-known classical thinker to emerge from the old ruling aristocracy. His class background surely helps explain his dogged refocusing this line of argument from the point of view of the feudal royalty. Another obvious explanation was the intended audience for how 法 fastandard works. However, a further plausible explanation for his choosing this way of putting the argument was his own philosophical training. Hanfeizi's use of these arguments form a royalist reaction to the position of the Confucian authoritarian, Xunzi.
Hanfeizi was a student of Xunzi, not of Shang Yang. Shang Yang, as far as we know, worked out his "political theory" in the course of actual practice of reforming and running a state.  Confucians worry that Xunzi is unfairly blamed for the advent of Legalism. He is not to blame in the sense that his doctrine is anti-royalist. However, his doctrine is clearly an plausible explanatory factor and he deserves blame for the content of that doctrine! Let us see why.
In the Xunzi, we find several lines of argument that seem like attempts to capitalize on the growing and obvious appeal of 法 fastandards. The original motivation of 法 fastandard theory, as we saw, was anti-Confucian. The Xunzi seeks to hijacked its critics' concept for his own Confucian purposes. Like the Guanzi, the Xunzi advocates both 法 fastandard and 禮 liritual (traditional rules of decorum). Xunzi's way of combining them, however, was to the opposite purpose. He did clearly treat fa as a standard of interpretation, not as anything like a law. However, Xunzi argued that the only standard or norm of correct use of words is standing convention. Thus, he concluded, all reforms were misuses. He classes the reformers with the sophists and their distortions of the correct use of words. The wise king would forbid such talk and punish anyone thus misusing normative terms in attempting to reform 道 daoguide.
Given the intended conclusion, i.e., traditional Confucian禮liritual are the correct 道daoguide, the Xunzi position must address the traditional Confucian question of what counts as the correct interpretation (performance) of these conventions. The Xunzi answers that the only viable 法 fastandard of correct performance is the model of scholars of that traditionthe Confucian junzi. We should validate current junzi by how previous junzi trained them. If, in the eyes of their teachers, they had learned, then they count as法 fastandard for the next. The chain of validation ideally goes back to the coiners of the traditions. This gives Xunzi a version of the claim that tradition is necessarily correct.  In treating the current cultivated intuition of the Confucian gentleman scholar, the junzi, as fa, he was rehabilitating the idea of the acquired, schooled intuition into the norms for the use of fa--the very thing with which Mozi and Guanzi originally contrasted fa. He appropriated his opponents' word and reversed its meaning.
Since the Xunzi shows little sign of having absorbed the Analects salutary aversion to punishment (in fact betrays as much enthusiasm for punishment as any good legalist would) the Hanfeizi diagnoses this Xunzi line of argument as a transparent power-play. The Hanfeizi analysis is that if the code governing behavior is the 禮 liritual, then indeed only the Confucian intelligentsia could correctly administer the normshence their attitudes would determine who was punished and rewarded. The king would become a figurehead and the scholarly class would be the effective rulers, imposing their intuitions about right and wrong with force.
The Hanfeizi and the Xunzi share an extreme preoccupation with what they perceive as the misuse of words. The Hanfeizi joined the Xunzi in encouraging the ruler to ban all moral reform (or linguistic reform) talk. However, he again put the point only in the context of how it might endanger a ruler. It does this when his high officials and the bureaucracy control him by flattery, displays of learning, moralizing and other persuasive trickery. To protect the ruler from this rhetorical manipulation, Hanfeizi argued that a royal should govern his own choices by objective, measurement-like 法fastandard .
If we forget the role of fa as objective standards for the use of a word, we can miss the point again and think that this is Hanfeizi arguing for rule of law--of law as governing the king. However, it is easy to see that Hanfeizi had no sense of any such value. He is a great storyteller and many of his fascinating anecdotes are celebrations of arbitrary punishment by the king. The king's concubine winces at his bad breath, so he immediately orders her nose cut off. He does this without any reference to a law he had passed forbidding halitosis hate-crimes. The story appeals to Hanfeizi simply as way to enhance fear of the ruler and encourage a habit of absolute, total obedience. Hanfeizi is an advocate of rule of officials by law, but that is as close as he gets to understanding the rule of law.
So Hanfeizi didn't give us much of an argument for rule of law but, as we've seen, the line of reasoning was widely available in ancient China. Controlling official action by predictable guidelines allows individuals to choose a course of life that reliably avoids the risk of punishment. Hanfeizi performed a service by selling the idea of a rule of fa to the ruler.
So what went wrong? Why would the culture accept a reversion to a Confucian rule of man?
The usual story cites three factors: a) the cruelty of legalist punishments, 2) the rigidity in interpretation of the penal codes and 3) the exponential growth in complication of these codes. The first certainly could have been a factor, but cruel punishments themselves originated long before the idea of 法 fastandard and mostly survived the Confucian counter-revolution. As we saw, Xunzi shows enthusiasm for the notorious mutilations equal to that of any Legalist. Legalists were, however, certainly consistent advocates of cruel punishments including punishments way out of proportion to the crime on the theory that if we have clear, explicit standards of performance, severe enough punishments would never actually be employed. So the rigidity of fa may lie behind their justification of exceptional cruelty.
The key problem, I believe, was the rigidity. The initially liberating idea of a measurement-like standard became an interpretive millstone when magistrates had to respond to the complexity of actual infractions and hard cases. The whole idea was to have standards so rigid they did not require or allow for intuitive notions of whether the punishments were right or wrong. The moral theory at the time did not say much about 'intent' nor did Confucians develop a theory of excuses, responsibility and mitigation. Excuses and responsibility were simply implicit in the appeal to intuition. This surely heightened the sense of injustice when legalists coupled "mechanical" enforcement with their extravagant and disproportionate punishments.
The rigidity also may be a culprit in the growing complexity of the codes. Every loophole, conflict, excuse and exception allowed in cases had to be corrected by writing a more complex public promulgation making the penal code ever more cumbersome. The result, as this insightful Confucian objection observes, is that the penal code loses the very value that justified it in the first place. The people no longer have predictable foreknowledge of what is likely to bring punishments on their heads. They do not know how to move hand or foot.
The cultural choice to blend in elements of a Confucian rule of man has to be understood as a choice among live options. In China, it was between a rule of law with an excessively rigid interpretive procedure or the live alternativeConfucian intuition. They had to choose between the acknowledged good of predictability and the attendant control of one's own life and the evil of rigidity, injustice and the ever more complex codes. The live alternative to rigid measurement-like standards was the humane Confucian moral intuition. They chose to temper the penal codes with a cultivated moral judgment--a choice that would only work if the 禮 liritual became part of the penal code.
Western interpreters tend to pose the issue in their own terms What China needed, they would say, was more emphasis on the "spirit" rather than the letter of the law. However, the choice to which this slogan points in the West was not a live option. We read the slogan against the background of our natural law ideology. It implicitly meant interpret the law morally. The "spirit" of Western law invoked the ideal of a fully rational norms uniting all realms of law.
In ancient China, the live alternative was the cultivated intuition. It was the closest analogue to the "spirit" of the code and was available, as we saw, only for the ritual code. Thinkers would have had little justification to believe in a "natural human intuition" into the point of the ruler's arbitrary commands! They handed the job of judging and punishing back to specialists in the Confucian ethical conceptions rather than in the law.
This retains the use of punishment to control people, but removes the means to control officials--objective standards of interpretation. So the "compromise" choice of the dynastic system up to modern times blended the expedient but morally dubious rule by law with the flexibility of the rule of man. What the modern consensus in Greater China amounts to is that this was, in hindsight, the worst of both worlds. It betrays the original Confucian argument against rule by law and abandons the widely appealing anti-Confucian idea of protecting people from arbitrary official authority.
The first point of this narrative is that the moral ideal of a rule of law is an authentic ancient Chinese value. It is a part of the Chinese ethical substance. An interesting sidelight is that it shows that we do not require the notion of "rights" or a moral individualism to justify the rule of law. The common moral outlook required was an authentically ancient Chinese spirit of equal concern for all. We saw this in the Mohists, who originated the notion of fa and advocated universal love. We also notice it in their skepticism of Confucian claims of special moral authority -- a skepticism expressed even more openly in Daoism and Legalism. Daoist emphasis on equality is even more radical--an equality based on rejecting any moral doctrine as absolute of treating everyone, even those whose toes and feet had been chopped off (probably as punishment) with equal respect. It is perhaps not accidental that history sees Daoist inspiration behind the recurrent waves of popular rebellion against intellectuals and elite rule in China--right down to the Cultural Revolution and the Falun Gong.
How, then, should we go about making this ideal with its deep roots in the Chinese ethical "substance" work this time? Western science offers a helpful techniqueinterpretation as an analogue of scientific theorizing rather than measurement. In the West, this method has become the common property of the realms linked by the shared concept of law.
Otherwise, the West has moved well past its ancient natural law foundations. The legal positivism movement effectively separated law and morality bringing the two views of law closer together. But before the three realms were separated, they had all worked out versions of a model of reasoning which gives them an alternative conception of the interpretive "spirit" of the law.
In science, we postulate laws to explain the past and to predict future observations. We search for a coherent set of laws to explain everything. In morals, similarly, we seek for "first principles" to justify our actions and intuitions. So, in law, Western theorists similarly use a technique of principled reasoning. When we have an intuition about how to decide a case, e.g., the free speech case, we should formulate a guiding legal principle that justifies our intuition and also justifies our past decisions. We feel bound to be consistent in our principles of interpretation. So we propose a principle, say the famous "clear and present danger" test. Then we test that proposed principle by considering its result in hypothetical future cases. Can we legally forbid shouting fire in a crowded theatre? If we find no acceptable principle that justifies our favored decision in this case, we must reconsider our intuitive judgment and try again.
We in Hong Kong can now see it as an advantage that this conception of the "spirit" of the law that it does not require linking it to a particular moral view. It means we do not have to adopt an official, comprehensive "morality." Society can entertain and tolerate its members following and advocating various directions of moral reform both. These different attitudes can co-exist and compete under a common and fair scheme of law. We can have a cooperative society and open and vigorous discussion about the right way of life. We expect our judges to be experts in legal principle but not advocates of any particular moral vision.
However complex it becomes in the face of economic and technological developments, the law seeks a rational, principled structure. This gives it a way to balance its two moral purposes (1) it control officials in their use of state force in ways that individuals can reasonably predict and (2) it can be used for social reform of existing normative practice. The principles are stable principles of interpretation of changeable statutes. Principled interpretation also allows interpretive flexibility. Law need not be rigid in the ways that led to the obvious and blatant injustice that caused Chinese culture to recoil from two millennia ago. Of course, as we all know, the laws predictability from the ordinary man's point of view is far from perfect. Thus, the shared Chinese and Western justification can not underwrite the policy of the ancient Legalists when they argued for disproportionate and cruel punishments. Since law is imperfectly consistent and imperfectly predictable, punishments must be both measured and mild.
The moral ideal of a rule by a system that ensures people a measure of protection against arbitrary official coercion is not a peculiarly Western value. The idea that such a system allows people to choose courses of action that are free from risk of punishment has deep roots in Chinese tradition -- even in Confucianism. However, its implementation is a difficult matter and is a place where we can borrow from Western interpretive techniques. The result should help all of us achieve a shared goal of Western, Confucian, Legalist and Daoist philosophers--avoid punishment!
 Keith (1994) argues that this consensus in China continues despite June 4.
 See Hansen (1985)
 See Peerenboom (1999:320) for citations of the discussion surrounding this distinction in describing China's attitude toward the role and justification of law.
 I do not, for example, treat it as essential to the distinction that rule by law places rulers are above the law and rule of law places law above the rulers. One who justified law in the way I characterize as rule by law could, for separate reasons (democratic inclinations, moral commitment to equality, practical worries about people's allegiance etc.) also advocate that rulers be subject to the law. The difference would be that this attitude would not be a consequence of the justification of legal institutions and practices. By contrast, a strict consequence of justification of the rule of law attitude is that rulers are bound by law.
 It should be obvious that, given the way I use the distinction, many popular Western European views about law would count as rule by law. Early and simple-minded legal positivism, for example, seems to accept this position. That is why realists could challenge them with the observation that if their theory of law were correct, we would have no reason for fidelity to law (no rational justification of the institution of law). See Fuller (1958).
 I mean logical, not historical priority here. The syntactic/semantic conception does not play its role mainly in understanding penal law but in the scientific and moral contexts where we postulate axiomatic structure.
 The definition of law was the main topic of philosophy of law until the latter half of the 20th century, even up to H. L. A. Hart. Now the explicitly normative focus has emerged from realist criticisms of positivism and Dworkin's analysis which embeds definitional questions in a broadly normative, ethical-political theory justifying the rule of law.
 St. Thomas Aquinas: Summa Theologica Excerpted in Feinberg and Gross, Philosophy of Law 4th Edition. (Belmont, CA, Wordsworth:Dickenson ) 1975 p. 14.
 See John Austin, "A Positive Conception of Law" from The Province of Jurispurdence Determined. Excepted in Feinberg and Gross (1975). This quotation from p. 26.
 See, for example, the discussions in Bodde (1981).
 Assume for these purposes that the nature of morality is still controversial, with traditionalist Confucians advocating a 禮liritual 義yimorality , Mohists a利 libenefit 義yimorality or a 仁renhumanity 義yimorality and so forth.
 Munro (1969:110) notes that the Zuozhuan dates publication events to 536, 513 and 501 BC.
 Bodde (1981) p. 171.
 It is notorious for its alleged graphic, possibly etymological and semantic connections to xingcarve and the heavy use of mutilations, amputations and tattooing as punishments in ancient China.
 It is tempting to see an implicit precursor of the Fajia matching of word and deed in the notion of the carving/tattooing marking or symbolizing the correct offense. We can also see how this conception could eventually be adapted for use in and argument for rule of lawby treating publication of the codes as limiting or regulating punishment.
 Robert Eno argues that Confucius' either had no political theory or that it was implicitly anti-government authority. See his 1990.
 Zichan's reply to the "Confucian" criticism of Shu Xiang hints at this line of defense. " I have neither the talents nor ability to act for posterigy. My object is to save the present age." See the account in Bodde (1981:178).
 Scholars frequently note other elements in other Confucian attacks on publication of the codes. See Munro (1969:110-12). Below I acknowledge but dismiss Confucian appeals to their own class interest as "superiors" and opposition to equal treatment under the law as obvious special pleading. Similarly, as I note here, the appeal to traditional norms simply begs the question in the context of the reform agenda promoted by the "Legalists." While agreeing that the Confucian argument is a unique instance of non-anarchist opposition to legal process and institutions, I do reject the implicit invitation in Bodde (ibid) to treat the opposition to legal codes as somehow more genuinely Chinese than the attitudes of the reformers. "What is uniquely Chinese and therefore most significant about the letter, however, is its insistence upon the moral and political dangers involved in the public promulgation of legal norms. This view of law seems to have no real parallel in any other civilization."
 Zuozhuan "Zhao Gong 29" translation from Bodde (ibid).
 See the famous account of Dengxi in the Lüshi Chunqiu recounted in Hsiao (1979) p. 368-9.
 Arguably, this recognition is in the Mencius on which the Song-Ming theorists relied. That however is controversial as an interpretive analysis among Mencius scholars. See the discussion between Shun (1991, 1997) and Ivanhoe (1990).
 I am here accepting the Confucian values point that Confucianism conservativism does conflict with the rule of law but not that Chinese values do. Confucian values are irrelevant to the cultural point. They are important, however, in coming to understand that the moral reformists in China were soundly blocked from plausibly advocating rule by law. However, as I will argue, the liberal argument for rule of law remains as compelling in a culturally Chinese conceptual context as it does in the Western contextthough the contexts are different.
 For details of the Mohist argument see Hansen (1989 and 1992:Ch.4).
 See treatment in Graham (1989:274) for insight into and the context of this claim.
 I recently discovered a clear statement of this way of reading the legalist movement in Chien Mu. This quotation is translated by Mote in Hsiao (1979:378). From the time when punishments came to be regularized by statutes, the rewards and punishments meted out to the common people were no longer totally dependent on the pleasure and anger of the aristocrats, but now were provided with a reference point on the basis of which challenges could be made. Teng Hsi Tzus Bamboo Penal Code appears to have served as the instrument for instructing the people how to make such challenges, and those in high position at that time had no choice but to appropriate that [instrument] by which such challenges arose, and employ it in their own governing. This constituted the crux of one of that ages great changes.
 Graham (1974:277-8) is typical in addressing this claim mainly by looking at Fajia arguments against moral suasion, but leaving the supposed mechanism by which 法 fastandard works to achieve this "kind" result quite mysterious. What makes it mysterious is his orthodox assumption that 法 fastandard means standards enforced by punishment (reiterated later as "harsh law") rather than objective, operational, measurement standards for punishment and reward. He is clear enough earlier that this is what 法 fastandard are, but when he comes to these kinds of legalist claims, the reversion to the orthodoxy about 法 fastandard meaning "harsh law" blocks the obvious line of explanation of how 法 fastandard can be beneficial to people.
 I address this meaning-change hypothesis at greater length in Hansen (1993).
 The scholarly consensus has been that the book bearing his name was almost certainly not written by Shang Yang. That is not too striking since, strictly speaking, it may be true of most ancient "Zimaster"texts. Current textual theory of Chinese "Zimaster" texts is that they were "maintained" by textual communities of adherents to the text. In the standard case, the first community should be students. In Shang Yang's case, the consensus would normally imply that no member of the textual community responsible for any part of the book was an actual student of Shang Yang.
 This line of thought is, I think, the most compelling and plausible in the Xunzi, but as may scholars argue, conflicts with another much more dogmatic and absolutist theme. See Yearly