Summary

This article describes a new approach toward discourse ethics, defined as a discourse theory of justice or the right law respectively. It is significantly different from the classic discourse theory of Karl-Otto Apel and Jürgen Habermas, to some extent also from Robert Alexy. Those differences concern (1) many aspects of demonstrating why objective or universal norms can exist at all and (2) the content of those norms, particularly of the principle of freedom including its intergenerational and global, i.e., sustainability oriented, extension. Altogether, it still seems true that discourse theory is the most promising approach toward a modern universalism and a modern law of reason. In any case, the controversy about theories of discourse rationality should bring about much clearer arguments than the conventional debate about “positivism or law of nature” (the very notion seems misleading), in favor of and against universalism. At the same time, the article criticizes some aspects of economic theories of efficiency – to be distinguished from effectiveness – which actually do not concern “another aspect besides justice” but rather describe a (wrong) ethics.

1. Terminology: Justice, Action, Normative Reason, Instrumental Reason

Under what circumstances can we call social life “just”, or the law “right”? This is the ultimate question of all thinking about politics, morals, and the law. Conceptually, the term justice is concerned with the normative validity of a society’s basic order. Thus, a normative theory of justice (or morality) answers the question: How shall humans behave or what shall the founding order look like? This must strictly be distinguished from the question of how humans factually do act and what the factual reasons for this action are (and what humans factually “deem right”) – this is a question of the descriptive action theory (or anthropology or the doctrine of the idea of man[2] – or theory of society). A link between the theory of justice and the action theory is the equally empirical governance theory or control theory, i.e., the doctrine of the choice of means to effectively and factually enforce previously defined normative aims (e.g., the right to freedom from impairments to life and health), possibly after a normative balancing with other conflicting objectives (e.g., economic freedom). Such means or instruments could be for instance taxes, voluntary commitments, regulatory law, or competition.

These statements are purely definitional and do not yet concern the content of, e.g., justice.[3] In our context, justice involves more than what many people associate with this concept: the correct distribution of material goods is only one aspect of that general justice (and may be called “social distributive justice”). The following clarifications are also important:

• “objective” means “not subjective”, i.e., independent of certain perspectives, cultural backgrounds, or preferences – thus universally and everywhere applicable.
• “Reason” or “rationality,” respectively, refers to the ability to decide questions objectively. When it comes to the question of the validity of moral and legal principles, norms, objectives, or judgments in general, including the question of how to balance among those aspects (!), we call this normative reason. In contrast, instrumental and theoretical reason concern facts, thus descriptive truth and not normative justice. Instrumental reason considers the means to implement a given objective which is accepted as correct, e.g., a specific climate target (or a very selfish goal like a theft), in a most effective way – e.g., through taxes or emissions trading. Theoretical reason deals with the finding of facts without a reference to a concrete action, such as in scientific climate research. Facts from theoretical reason are used in the course of balancing under normative reason.

Whether there are objectively valid (i.e., rationally provable) norms and facts, has nothing to do with the correct observation that as human beings our knowledge about facts and norms is often factually impaired by our subjective points of view. This tendency to a subjective perception is natural. This, however, by no means proves that objectivity – for example through careful examination and discussion with others – is simply impossible.[4] Let me give an example: It may be the case that there are scientists who express their opinions for or against the existence of human-induced climate change because they expect economic benefits for themselves from those actions. Therefore, their statements were not objective, but subjectively distorted. However, this in no way means that there cannot be objective and unbiased knowledge on climate change. Rather finding factually very “subjective” perspectives logically requires that there are objective perspectives – otherwise the subjective element of those subjective perspectives could not reasonably be determined. However, most economists, sociologists, and political scientists openly deny the possibility of objective statements regarding normative questions (they only accept objective statements regarding factual questions[5]). The following will analyze this position.

2. Why skeptical, contextualist, religious, etc. theories of justice are problematic – and why Rawls and Habermas are also insufficient

Why is only that basic order just, which promises freedom and democracy, be it in Europe, in France, in India, in Germany, or anywhere else? Philosophical liberals[6] are of the opinion that such an order is universally and exclusively true. Their argument is usually based on human dignity and the principle of impartiality. Dignity (equal respect) is used here as the necessary respect for the individual as an autonomous being; impartiality (in spiritual succession and clarification of Kant’s categorical imperative) shall be regarded as a principle of not asserting special interests. The same respect and impartiality are brought together in a memorable picture, the “veil of ignorance”, by the most famous Kantian of the 20th Century, John Rawls. In this hypothetical situation, called original position, equal decision makers face each other under the condition that they do not know who they are going to be in real life. This picture illustrates the two basic principles but does not explain their existence. Furthermore, it does not answer the follow-up question of the skeptic, who understands justice as any arbitrary subjective or social construction (or of the contextualist, who simply assumes that the long-grown moral[7] tradition of his culture is right): why dignity and impartiality? And this question is important because skeptical and contextualist positions are increasingly dominating the arts section, politics, philosophy, political science, and jurisprudence[8] – and they are characterized by the fact that they deem judgments and questions of justice either not decidable at all (skeptics) or decidable only based on our respective non-universal “cultural values” or “factual everyday moral intuitions” (contextualists).

Rawls noticed the lack of justification in his two liberal core principles of respect and impartiality. He therefore offers to those to whom these principles do not appear self-evident as “justification” a reflective equilibrium: This elicits, quasi empirically, our two central moral intuitions. As Rawls himself observes “the hypothetical nature of the original position invites the question: why should we take any interest in it, moral or otherwise? Recall the answer: the conditions embodied in the description of this situation are ones that we do in fact accept.”[9] According to Rawls, respect and impartiality result from our de facto recognition of those basic principles. Unlike real contextualists[10], Rawls‘ “minimum contextualism,” does not simply declare “all our everyday moral intuitions“ just. The influence of intuition in his theory is limited to the two principles, which then are the basis for further deductions. Still this problem prohibits Rawls from offering a truly universalistic theory since the “two principles” are probably not factually recognized worldwide.

Moreover, Rawls fails on certain issues, which also contextualists or communitarians stumble upon: (1) Why should certain “factual intuitions,” i.e., norms that are factually recognized, therefore be deemed right? (2) And whose intuitions are decisive: those of the majority, those of the general public, or of a two-thirds majority? And if the majority: Why exactly the majority? (3) How can all this be applied without resulting in the is-ought problem (which arises where factual intuitions/ judgments are used to determine their validity)?[11] (4) And how can deriving judgments from “our factual everyday intuitions” ever give solutions for new normative problems such as the issues of intergenerational and global justice, which require breaking up old traditions? One last point: (5) How does a contextualist handle the problem, that his approach necessarily approves a deliberate totalitarian dictatorship of the majority? (Unless he first introduces a particular non-totalitarian moral ideal. Then again, we would have to ask why just this ideal should be binding – this can be expected to lead either into a dogmatic setting without justification or an infinite regress).

All these problems also occur if a contextualistic approach states: “Right, is exactly that basic order which complies with our moral sentiments.” At first, this sounds interesting. Why would you bother to rationally explain, e.g., that National Socialism was unjust, if we already have an “infallible feeling” that “a Nazi is a swine”? But this is simply not enough. Otherwise I would just as well be right if I happened to have an “infallible feeling” that “all Jews should be killed”. And what are you going to do if you encounter a skinhead who actually has just this sentiment? Relying only on feelings could not justify why the “nice” should somehow be right, and the “evil” be wrong. Therefore, by definition, no universal standards of justice could exist. Kant, unlike Rawls, searched for an answer to the question of justice, which would be free from empirical assumptions (and universally applicable) in order to avoid, inter alia, such critical, ultimately unsolvable objections: To Kant precisely that basic order is just which arises from the normative practical reason. Unfortunately, he cannot show convincingly why (a) dignity (respect) and impartiality are “necessarily” implied in normative reason, and why (b) reason is (even universally) inescapable.[12]

A skeptic to a theory of justice could object: “How true! All attempts of founding a concept of just societies fail. Consequently, justice, and generally any judgment, is always irrational and therefore purely subjective.” This is an enhanced version of the ever more popular – conservative or left – position, which already the ancient Sophists used against Plato. In this sense, skeptics assume that it is unthinkable to justify norms of judgments. Sometimes this is coupled with utilitarian or Hobbesian features, where at least normative balancing of the “infallible” factual preferences of the citizens is considered rationally feasible – if such balancing is not considered a mathematical statement of fact based on clearable preferences. Note that this also addresses the standard approach in economics, which receives increasing attention in political debates. Therefore, these ideas require further analysis.

However, skeptical approaches already stumble upon a core problem, which is also an issue for contextualism (besides those frictions, which have already been presented). First: A contextualist’s basic idea that all normative statements depend on their cultural context is logically incongruent, since he necessarily uses as a background assumption what he denies at the same time. Because the statement itself, that what is derived from a cultural context is per se right, cannot meaningfully be understood as context dependent in the sense of: “We factually think that what factually can be found is right.” For empirically this would be wrong since many residents of the West, and of the entire world, factually do not make their judgments based on a cultural context, but, e.g., on religion or liberal ideas. Anyone who states: “Everything is based on context”, therefore more likely means that at least this statement itself is independent of its context or any culture. In doing this, however, the contextualist applies what he considers impossible with respect to liberal ideas: He relies on a universal level, where people can gain insights, which are true cross-culturally. Ergo he relies on reason. Thus, reason, which by no means depends on the respective culture, is the ultimate and universal instance, which is the starting point for discussions about justice (not the social context). The skeptic is caught in the same trouble: If his statement that everything is contingent, purely subjective, and depends on the respective observer should be correct, it must be binding on itself. For if any norm is only an unjustifiable, not further rationalized “subjective construction,” obviously skepticism itself is only such a subjective construction of the world. Then, however, it cannot rule out objectivity / universality. On the other hand, if the skeptical thesis wants to claim its general validity, it contradicts and thus cancels itself out – because then there are obviously not only “subjective opinions” but also universally correct statements. Then, however, the liberal conception, which considers issues of justice to be rationally decidable, is not disproved. This is also true if the skeptic calls judgments unjustifiable but at the same time tries to justify the underlying judgment[13] of this claim (e.g.: “Norms are unjustifiable because there are obviously different systems of norms around the world.”). Hence, reason is a basis for determining issues of justice, which cannot be proved nonexistent.[14]

Laissez-faire and utilitarian preference theories, which are the common theories of justice among economists and partly among sociologists and political scientists[15], already fail due to this contradiction that proves normative reason inescapable. It is a mild variant of a liberal-positivist skepticism subsequent to Hobbes. The basic thesis is that that order is just, which would result if one solely used the (usually selfish) factual preferences of citizens and (optimally) translated them into clearable monetary units.[16] Any preference theorist considers the factual wishes of the people per se not subject to any criticism. The utilitarian then uses the highest possible sum of the factual preferences of the citizens as a criterion of justice (“the greatest happiness for the greatest number”). The Hobbesian or laissez-faire economist, on the other hand, looks to a consensus of selfish people. And in political science majority theories, the relevant preference is what the majority of citizens factually decides (possibly under a specific procedure. However, how such a procedure is based on a rational theory usually remains unclear). All this primarily uses instrumental (empirical) rationality, while norms or judgments are declared irrational in their basic principles. After all, preference theorists – unlike other skeptics – recognize a rational balancing between those preferences (yet such balancing is not understood as a normative act but rather as a quantifying clearing of divergent preferences). Moreover, unlike real skeptic, those theorists may consider the whole process as a kind of “justification”.

However, all preference theorists must fail due to several objections and questions that ultimately continue my criticism on contextualism: (1) Why should factual preferences be right per se? Was the Third Reich “right” due to its de facto approval of many citizens? Since even here it seems impossible to give an answer, preference theories end in infinite regress, or they rely on a dogmatic setting – and thus are unjustified. (2) But most of all, the thesis underlying the orientation toward empirical preferences (or “needs”) that there was no normative rationality or no justified goals, cancels itself out, as was demonstrated with respect to skeptical theories. (3) Furthermore, a preference theory is not only powerless regarding dictatorship, but also (4) when faced with new challenges (e.g., in terms of intergenerational and global justice) and new guidelines are seemingly too uncomfortable, unfamiliar, or too expensive. (5) Furthermore, a mathematical calculation or quantification of all preferences is not reasonably possible, for instance because they very often do not have a market value.[17] (6) Finally, an inference from preferences to their normative validity encounters the is-ought problem. Despite all these faults, I share the basic intention of many economists that every person has a right to his own, autonomous pursuit of happiness, which no religion, no cultural tradition, no authority, etc., may bar. What I mean is this: every citizen has to decide about his good life and society is only concerned with issues of justice – this thesis, however, can be justified otherwise without the demonstrated faults (namely through freedom rights, which will be justified in the following by transcendentally relating them back to our previous findings). Thus normative reason as the basis of justice cannot be proved impossible. Incidentally, this shows that the alleged antitheses “efficiency versus justice” or “efficiency versus ethics” which economists and their leftist critics like to use make very little sense. In fact, the preference theory as the core of economic efficiency theories is itself an ethics – just not a convincing one.

But what takes the place of preference theories? The most stimulating recent attempt to justify freedom and democracy (perhaps universally) comes from the discourse ethics. Jürgen Habermas, Karl-Otto Apel, Robert Alexy and others no longer understand normative reason like Kant as something of substantive nature, but simply as the capacity to decide normative questions based on reasons. Habermas, for instance[18], then formulates the principle of impartiality, i.e., the principle that the social order must be generally acceptable independent of special interests, as a central principle of justice. But unlike Kant, he does not only postulate it. Rather, he founds it on the inevitability of reason which was just shown and on the following considerations: judgments and knowledge in general is always tied to the medium of language and thus to “the understanding” among men. Since such an understanding could only be achieved through reasoning, we would always have to look for a consensus in discussions about justice. And this is precisely the principle of impartiality: namely, that exactly that norm, and that order, is right, which all possible interested parties could agree to without duress. In addition, everyone had to respect its dialogue partners as equals. Unlike Kant, however, Habermas seems to understand the idea of mutual respect particularly for the autonomous individual (which is separate from the respect we have just talked about and which is the foundation of freedom and democracy) as depending on the respective cultural context. In his view, it is a mere “formal property of modern law” and thus somewhat more occidental than universal.

This classical discourse ethics, however, is again subject to important desiderata.[19] First, it encounters the same difficulties we have already seen in our discussion about Rawls with respect to the dependence of some aspects (“formal properties”) on their cultural context. Second, most discourse ethicists use additional vague background assumptions (that each allegation a person makes contains an implicit claim of validity; that the need to look for consensus is a feature of language; that “discourse obligation” and a duty to rationality, which substitutes reason for man as the ultimate goal), which I will dispute in the following. Third, particularly Habermas and Alexy (in more detail below) do not sufficiently clarify a very important point: that not only rules for our discussions about justice but also for actions in justice may be derived directly from normative reason, and that there are obligatory reasons for this assumption. Fourth, the classical discourse ethics has so far only provided some elements but not a comprehensive theory of justice. For instance, there is no theory of freedom based on discourse ethics (certainly not one that is universal in all societies, intertemporal between times, and global between societies[20]) – and therefore, previous discourse ethics may lack the core of all justice. Furthermore, there is no clarification whether human reason makes the autonomous individual the relevant standard, which is essential in the dispute about universalism. Fifth, the classical discourse ethics (see below for details) has not yet recognized that a final concept of a just basic order can rationally be deducted. Such a concept is not only comprehensive but also does not allow any additional principles – which could then qualify, e.g., freedom and democracy.

3. The case for a more advanced universalist discourse theory

Against this background, I would like to propose a modified rational universalism[21] or an advanced discourse ethics, which I consider more persuasive and not subject to the previously discussed shortcomings. This essay, however, is limited to introducing the rational basis of justice – it cannot explain in detail the immediate implications for the theories of freedom, democracy, balancing, balance of powers, and control, or the extension of all these theories in intertemporal and global perspectives (sustainability). This new approach breaks with old liberal-Kantian, but also other concepts in terms of their rational foundation particularly by offering a transcendental (or elenctic) argument which Kant probably considered impossible with respect to practical philosophy at the end of his lifetime. Transcendental – respectively negative or elenctic – arguments show that a certain statement (e.g., a specific norm) cannot be denied without presuming the very norm.[22] Ultimately, I want to show: (a) a basic order is universally reasonable, and thus just, if it is based on dignity, impartiality and freedom (and operates on rules of balancing and institutional rules (democracy and balance of powers etc.) which are derived from those very principles. (b) Reason as the basis of justice is itself inescapable.

I start with the argument of the “regular” (not strictly logically “transcendental”) lack of alternatives to (b) reason or (a) the liberal principles of respect and impartiality. As we have just seen, the deduction of principles of justice of a basic order from a religion, from a particular cultural tradition, or from selfish preferences has failed. That does not only mean that non-liberal theories of justice are untenable. It also follows that there does not seem to be an alternative to reasoning and discussing what is just. In other words: there is no other way to justice than through reason. But if the basic order may be reasonable and since there are no alternatives – due to the refutation of the other theories – even must be reasonable (as I will reinforce below): What principles of a basic order deserve to be called “justified through reason?” For “justified through reason with respect to what ought to be” is the (shallow) meaning of “normatively rational”. Contrary to the classical Enlightenment rationalism (and in a way contrary to Kant in his “Metaphysics of Morals”), tying the basic order to reason does not lead to a substantive list of principles. Therefore, we can say that normative reason does not have a specific content and is “open” in this respect. This openness, however, does by no means imply that the skeptic’s thesis of the arbitrariness of normative statements is correct. In fact, the previously shown possibility and the openness of reason imply a different conclusion: If no one knows who might give the best arguments in a dispute about justice, but if, at the same time, reason is possible and seems to be without an alternative, then we must assume that anyone who possesses some kind of reason, and thus every human being, might know the best arguments. Then there is no alternative to a basic order which allows such a dispute with reasons (this does not imply a duty to be reasonable at all times, as will be shown in the following). Therefore, principles of justice must be generally acceptable and ergo impartial, and we must respect our partners as equals. Hence not only is reason itself without an alternative, and thus not only possible but also necessary; the two principles of respect and impartiality are required as well. Thus, just like reason itself, the two liberal core principles are founded on a general lack of alternatives – therefore, they need not rely solely on arguments against other theories. (Later on, we will see that this argument is also valid beyond the discourse with and action toward a current dialog partner.)

This argument, that respect and impartiality are without an alternative, can be supported by a transcendental argument in favor of these two principles. The transcendental argument supporting the liberal principles can best be demonstrated by the following example (based on Robert Alexy[23]) which clarifies the logical implication of using reason with respect to issues of justice: No one could say “My argument X and the reasons could easily be refuted by Mr P, but you, Q, should believe it, stupid.” And no one could say: “After P had been silenced, we could finally agree that X is a good reason for Y.” Consequently, understanding the act of giving reasons as hierarchical or relative to the person of the addressee contradicts the very meaning of “reasons” – a good reason is convincing, and can be accepted by anyone. It follows that someone who, on the one hand, uses reasons in a conversation about justice (i.e., by arguing “because,” “therefore,” “since,” etc.), but on the other hand refuses to respect her dialog partner, acts contradictory since she denies what her own words (talking in reasons) logically imply under the rules of discourse. It follows that anyone who argues in a dispute about justice and thus engages in reason must respect her dialog partner as equal. This is true regardless of whether she is aware of the implications of her reasoning or assumes to be reasoning for purposes of persuasion only as we are considering strictly logical implications of speech (not a purely factual self-image, which in and of itself does not imply anything due to the is-ought problem). This respect for autonomy as self-determination, which reason requires, must be directed at the individual and thus be respect for the individual autonomy: for collectives as such cannot be possible discourse partners. It is rather the individual, arguing human being.[24] Thus, we have a clear foundation of the individualistic principle of respect and human dignity, which Kant still lacked, without having to rely on a de facto substantive idea of reason. In the same way, we can also derive the principle of impartiality. For the reasoned dispute about justice, which is without alternatives, must be generally acceptable and thus impartial with respect to its process and result because, again, there are no alternative substantive standards and because the category “reason” aims at equal and free appreciation. All of this is the foundation for liberal democracy – and it is universal, because it is tied to the human practice of speaking in reasons which transcends all cultural boundaries. Contrary to Habermas and Alexy, however, all this is not yet implied in our speaking. It rather follows from our reasoning.

One could argue: Personally, I will never speak to most people in the world. To see why this objection does not matter a further step for a liberal universalism is essential: not only my current discourse partner may claim respect and impartiality. If someone ever speaks in reason, this brings about the principle of respect and impartiality along with the subsequent broad liberties (freedom of speech, property, freedom of assembly, general freedom of action, etc.) even toward potential discourse partners, i.e., people we do not currently talk to. For reasons concerning issues of justice (unlike statements in purely private matters of aesthetics) are apparently aimed at anyone who could potentially refute them – thus I must respect all people once I have opened the discourse in reason. The following consideration can confirm this finding: No one could seriously say, “The absent Mr. P could easily refute my theses at any time – but you should believe them because of your stupidity.” This does not justify anything. It clarifies again the logical implications of the category “reason” – this time beyond the circle of current discourse partners. Solely relevant is again what „reason“ logically implies – not what I factually think or intend when speaking and arguing (because this factual assumption is normatively irrelevant due to the is-ought problem).

Dignity (or respect) and impartiality as the central discourse principles become comprehensively obligatory through the next step: they are necessarily principles of action. For the openness of reason renders it inevitable to keep the further discussion about a certain question (or a new discourse on another issue) open in principle. Who violates the principle of respect, even outside a discourse, limits the ability of other discourses, since any action (e.g., a law or an individual act) sets the framework for any further discourse. This thesis about discourse and action is again not only based on the argument that substantive reason, i.e., reason containing substantive principles, does not exist – an argument of a lack of alternatives that we have just inferred in two steps (= reason is open; and certain actions limit the possibilities of further discourses and, therefore, should be avoided in light of the idea of open reason). Likewise, it is not only supported by the inconsistency of anti-liberal models. In addition, the substance of the term “reasoning” which can be elicited through another transcendental argument supplements my thesis. No one could seriously say: “Although I know that my arguments would not convince you, if you thought about them for a week, you have to believe them today.” Hence, dignity and impartiality as a universal basis of just social life necessarily apply to discourses with respect to procedure a n d results. The gist of the matter: the combination of the action and the potentiality element of “reasoning” creates a normative space that binds a person even if she is not currently engaged in a discourse.[25] This is the foundation of a universal liberalism in the form of a justification of the principle of respect and impartiality, including its implications, i.e., the principles of freedom and democracy and their further consequences.

This required full autonomy and impartiality logically implies an extensive right to freedom in the sense of “absence of coercion”. For only with such a law, I can autonomously decide my life plans – a choice which may not be impaired in favor of special perspectives. But how can we justify the inevitability of reason itself? I already mentioned at the beginning of this section that it is (I) without alternatives. Furthermore, it was (II) mentioned that is at least unclear whether the proponent who defends universalism must bear the burden of proving his thesis. Moreover, and most importantly, (III) I have pointed out the existence of a transcendental argument in favor of reason. For who calls reasoning impossible (as ending in dogmatic settings), but nevertheless uses the category of reason herself by arguing in favor of reasoning being unjustifiable, and thus engages in reasoning on an issue of normative judgment – and, in addition, uses the category of reason regarding normative judgments elsewhere in life[26] – contradicts herself and cancels out her argument. It is well known that the same holds true for those who deny the existence of universally valid judgments altogether. For either such a statement is purely subjective (and does not refute anything) or, if it is objective, it presumes the existence of valid judgments itself. As was mentioned before, a statement on the question whether judgments exist is itself a statement in the field of normative issues and not an observable fact. If the skeptic argues that this question concerns a “meta-level” and was therefore a matter of fact, this would be an arbitrary dogmatic assertion. Finally, (IV) we simply presume the possibility of reason every day, whenever we establish normative assumptions and justify them, i.e., claim their objective acceptability (rather than presenting them as purely subjective). And it seems nearly impossible to refrain from using “because, since, therefore” on normative issues for a lifetime. Thus, the general possibility (!) of objectivity in normative questions is inescapable.

Therefore, reason and universal valid norms are possible. In addition, a rational basic order is necessary because (1) it is implied by the principle of impartiality. For only through reasoning we can clarify what is generally acceptable, and only in a rational basic order arguments can reasonably be analyzed. Furthermore, (2) rationality is implied by the fact that a universal level of validity is undeniable, but at the same time every other idea of justice fails. In short, the liberal principles already follow from our occasional use of reason (along with the possibility of reason) by the implications of reasoning. At the same time, these principles call for a constant use of reason (i.e., the necessity of reason). Yet for the present theory it would even suffice if reason was only possible but not necessary, since the liberal principles already follow from this possibility. Furthermore, reason need only prevail in questions of justice, where our once expressed arguments generate the liberal principles of justice, which in turn demand rationality in issues of justice. However, it is not required in questions of good life – for “reasoning” does not imply a generally binding aesthetics or way of happiness. Those remain our free choice (and vice versa, aesthetic discourses do not imply anything universal or even something that could displace justice). Still, the foregoing does not mean a duty to take part in a discourse. For our freedom, our autonomy and our discourses themselves would suffer from such an obligation to reason.

This entire theoretical approach cannot be doubted by rejecting the principle of consistency itself (which is in fact the logical basis of transcendental arguments). For consistency and thus logic (i.e., the possibility of inferring B from A, which is based on the idea that within a system of statements contradictions are not allowed) cannot be denied without self-contradiction. Because if someone says: “If I deny the imperative of consistency, I am not tied to liberalism”, then this person draws a conclusion herself – which is only possible if she presumes the validity of (classical) logic. [27] Besides, you can see here (not that this would be my topic in this article), that the foundations of classical logic, as such, are probably difficult to deny. For man necessarily uses the language of logic to be man. Furthermore, this language is not imposed on anyone. It is rather inescapable to everyone who does not want a lifetime of talking without reasons.

Thus, a space of universal principles is created from where we cannot escape – without the requirement of context or culture specific additions, without dogmatic settings, without a separation of action from the discourse, without a “duty to the discourse”, but with an emphasis on maximum autonomy of every human being. And even if we were silent from now on, we were still caught in universalism. For the unity of discourse and action has already thrown its inescapable ties on us which bind us with respect to all potential discourse partners. Thus, even if we immediately stopped to use reasons this would not change anything. Moreover, simply saying, “From now on silence,” is de facto impossible for man as man. Therefore, every day we renew our commitment to reason, respect, impartiality and freedom as soon as we speak. A fortiori, it is (practically) impossible to refrain from using the medium “reasoning” for life. And even through life-long abstinence of reasoning, one could reject liberalism only if I bear the burden of proof for justice and not the skeptic (which is by no means clear). In all this, we humans create this space of normativity. And since we are humans it cannot be excluded that we might err on everything which has been said so far. Because of this fallibility (which does not mean “subjectivity”) I am reluctant to use the term “ultimate justification” regarding any ethical approach.

Unlike the classical discourse ethics, this justification for respect, impartiality, and reason itself does not only imply certain liberties that are particularly evidently relevant for a discourse (such as freedom of speech, freedom of assembly, or freedom of association). Rather, the inevitability of the respect for autonomy requires a guarantee of freedom for all people. This includes every conceivable freedom that we know from the canon of human rights tradition. Only this grants full respect for autonomy – in terms of absence of heteronomy as complete as possible. For instance, whose freedom of occupation, secrecy of correspondence, property freedom of the press, or freedom to move is impaired, has fewer options to inform herself, to develop her own personality, to meet other people – and thus would ultimately be less able to face our arguments with her own free conviction. Who did not grant maximum freedom to his fellow, would therefore not fully respect her autonomy, since such restriction would be more than necessary to ensure the autonomy of all other people. Moreover, such a basic order would not be impartial because it would prefer certain plans of life to others. In passing, I note that such freedom is also attractive from the point of view of instrumental rationality respectively selfish interest. A justification for justice, however, does not require instrumental reason (which is not capable of justifying anything, as we saw with preference theories). On the other hand, the actual enforcement of justice heavily depends on instrumental rationality. But this goes beyond the scope of this article, and is indeed a completely different question: for the enforceability of a rule does not affect its justification. Otherwise we would have to justify the general proposition: “A rule that is in fact infringed upon, is therefore invalid”. And this sentence is clearly not justifiable (see below for an exception). Moreover, such a sentence would face the is-ought problem – the deduction of normative (in-)validity from a factual observation (low enforceability). Thus, contrary to the economic tradition but also to Alexy[28], a justification of freedom and democracy (the latter, implying a separation of powers and thus supporting impartiality and rationality, follows from the principle of respect and impartiality as well) does not require instrumental reason.

4. Refutation of some conceivable objections

The argument for reason and liberal principles that there are no alternatives (in its two variations: its transcendental and its regular form; not to mention the other argument that other theories are inconsistent or unconvincing) cannot be refuted by classical and hypothetical objections, as will be shown below. Altogether, it still seems true that discourse theory is the most promising approach toward a modern universalism and a modern law of reason. In any case, the controversy about theories of discourse rationality should bring about much clearer arguments in favor of and against universalism than the conventional debate about “positivism or law of nature” (the very notion seems misleading) does. In particular:

(4.1) First, it is harmless that even under maximum equal freedom as a prerequisite for discourses there will never be completely ideal discourses, which are free from any material, temporal or intellectual restrictions. For the impossible cannot be demanded normatively because norms are intended to resolve disputes – and an unreal order does not comply with this requirement. That is why our basic order need not enable absolutely ideal discourses, but only facilitate discourses which are as ideal as possible. And even these “non-ideal” discourses produce the liberal principles by means of logical implication, as we have seen. These principles are the normative framework for the possible results of discussing normative questions in politics, morals, or law.

(4.2) Second, my argument is not circular in such a way that I arbitrarily defined the meaning of the category “reasoning” using only supposedly universal significance, and thus in fact improperly universalized a typical occidental idea. For even a member of a completely “non-Western” culture such as a Trobriander engages in the category “reasoning” when he uses sentences with “because, since, therefore”, etc. It is in fact a human universal – even the proverbial Trobriander is therefore bound to the implication of this category, which she has brought about herself. Unlike Alexy and Apel (who build on the character of language and “claims” and not immediately on the “reasoning”, even though the former seems to imply very little) I do not hide my conclusions in a perverse use of the category “reasoning.” Nor do I use a list of good and less good reasons. I only assume sentences with words like “because”. Thus, my thoughts claim to be a simple reconstruction of what every man must logically presume, if he lives as a human being.[29] Even the Trobriander is not immune to this. It is clear, however, that enforcing a discourse rational universalism among “primitive peoples” could be difficult.

(4.3) Furthermore, this universalism in issues of justice is not cultural imperialism by making human rights and liberties a universal requirement of any basic order: First, very often human rights or liberties, which indeed fully determine the extent of just politics (i.e., freedom cannot be supplemented by additional principles unless those are derived from the basic principles as well; see below 5.) collide with each other. This generates large leeway for discourse in which a liberal theory only determines the rules of balancing (including democracy based on the balance of powers as the central procedural and institutional rule), but does not dictate a certain result. Second, guarantees of freedom create a space beyond the collisions of liberties in which everyone is free to act completely arbitrarily and may live according to any cultural or collective way of life. For example, if some Turkish women like to submit to their husbands, of course they can – it is only prohibited for the husband or father, e.g., to force his daughter to marry someone or, generally, to forcibly prevent the daughter/ wife from doing or compel her to do something. A general demand for “more tolerance with non-liberal cultures” would certainly be inappropriate. Involuntary widow burning in India etc. must not be respected as a “cultural tradition”. A theocracy, which can only be sustained by force, must equally be condemned because the respect for the individual is transcendentally required and religious or contextualist ideas are untenable.

(4.4) Furthermore, it is no valid objection that a discourse rational justice created a dictatorship of reason by referring to abstract criteria such as the possibility of general acceptance, autonomy and freedom. This follows directly from these principles and the is-ought distinction. For, first, we all may (in this respect Kant and Habermas are wrong) comply with the normative rational principles of justice for other non-normative rational motives – as long as our behavior could also be justified normatively rationally! The prohibition of homicide, e.g., is probably a normatively rationally required provision. If I comply with this ban, I am doing justice – even then, if I abide by the ban, for instance, due to sheer conformity or instrumentally rational fear of sanction. That therefore citizens’ “attitude” is a mere (empirical, associated with the control level) issue of enforcement and not a question of justice follows, again, from the fact that norms of justice are intended to solve conflicts and that consequently it must be possible to make them generally binding. At least the latter does not apply to attitudes. Second, we may not only think irrationally – regarding the good life, we may also act completely irrationally (and biased in the sense of “depending on special perspectives”). In fact, we do not a have a real choice anyway because there are simply no normatively rational and impartial answers to whether people in general should see a theater play or should rather go to the movies. A “midway” between the basic principles of justice and the private matters is every political, legal and philosophical balancing. It deals with the conflicting freedoms or conditions of freedom and is normatively rational, but give leeway. Third, – and crucial – no one is “heteronomously dictated” anything. I rather identify what conditions man cannot escape, if he does not leave the human form of existence as such. Unlike classical discourse ethics, autonomy as an intrinsic value is not despised estimated low. To the contrary, it is first and foremost justified. Fourth, anyone who criticizes that the liberal discursive basic principles are independent of the will of the majority would have to show conclusively why instead a system that “we factually agree” would be just. Was the Third Reich just through majority approval? (See above) Fifth, a discourse rational approach remains fallible, possible to criticize, and open to new insights. Of course, fallibility is not subjectivity or arbitrariness. We are not talking about (indeed arbitrary) definitions of terms, but about (justifiable) content. Regarding such content, we might find errors one day, as in natural sciences. Still, we could not call such knowledge arbitrary.

This does not change, if the hypothetical critic demands that we should at least be able explicitly to disengage ourselves from the basic liberal norms, as the logical implications of our own speech could not be stronger than our explicit statement. After all, this objection is irrelevant: first, we are not talking about “any implication”, but about something that I reaffirm every day through talking in reasons, which is inevitable for a human being. Moreover, secondly, this implication is logically inescapable. Contrary, the idea that one could “explicitly disengage oneself” is an arbitrary dogmatic assertion, which is ultimately rooted in the mere assumption of a general right for complete self-determination. And such a right as a mere assumption (as with Hobbes) is just not convincing. It is only convincing if based on a discourse rational reasoning – which leads us back to our implications.

(4.5) Moreover, you cannot argue that a discourse rational concept fails because there was more than one reasonable basic order, i.e., “many different rationalities”. E.g.: “It is rational for company A to set free as many workers as possible because has to make a profit – but it is not rational for the economy as a whole, because it increases poverty in general.” The following is true: Of course, the instrumental rationality of an action might be considered very differently, depending on aims. In our example, the actors, enterprise and state, have divergent goals, and we are considering instrumental rationality. However, “many different rationalities” only exist if one assumes that there is no normative rationality, so that objectives, norms and judgments as such are arbitrary and not rationally verifiable – and precisely this is not true for the normative basic order, as was demonstrated above.

(4.6) In addition, my approach does not encounter an is-ought problem by founding its normative theory on the existence of the category of “reasoning” in humans. However, it is formally true that such existence is a factual matter – unlike the implications of reasoning, namely respect and impartiality, which are logically derived and not empirically found by asking the speaker for his factual preferences). Nevertheless, a mere fact is indeed relevant for normative theories, if it identifies an empirical impossibility.[30] For instance, the norm “You all shall go jogging to Mars every morning” is not only problematic in terms of enforceability, but already regarding its justification since it definitely cannot be complied with. Because norms are intended to resolve conflicts; and if norms are not just difficult but impossible to comply with they do not resolve any conflict (even if there were any abstract good in regularly jogging to Mars). Therefore (empirically) impossible norms are not only unenforceable, but also wrong (which is why the norm that we should not resolve any conflicts is likely wrong as well). However, it is just not any simple fact that people talk in reasons. It does not actually occur that a human never talks in reasons – and it might be existentially impossible for people to do this before.[31] Thus there is no is-ought problem in my approach because is based on a theory of impossibility (though it is not a thesis of logical impossibility).

(4.7) A universalist liberalism cannot be successfully criticized by saying: “De facto, freedom, democracy, etc. are an invention of the West.” This tends to be true historically and sociologically (or with respect to anthropology) though not in every respect.[32] However, a sociological proof of origin does not show that liberal ideas are not normatively rationally valid. For what relevance should a purely factual statement about the origin of a norm have for the question of the validity of that norm (if we want to avoid an is-ought problem)? Ultimately, this is a confusion of factual genesis and normative validity. Genesis arguments as a critique of ideology can always lead to further verification of the validity of the norm – but logically they can never refute norms.

(4.8) This emphasis on the normative independence of a universalistic liberalism from “empiricism” and the respective “context” does not render my discourse ethics approach “context blind” – though some might think so at first glance. Of course, a normative universalism does not state: “In Taiwan, some women factually do serve their husbands. Therefore it is a good thing.” But first, we are free (see above) to use the margins of balancing where freedoms collide as well as our private space of the “good life” in order to live according to certain cultural traditions or certain “social contexts”. Second, though I do not admit facts – in addition to social facts natural facts are often included in the notion of “context” – as a reason for norms, they are indeed relevant for the application of the norm. We have already seen: While I cannot derive the prohibition of manslaughter (norm) from the fact that, for example, X killed Y yesterday (fact). The actual killing is still relevant when I have to apply a norm, here “prohibition of manslaughter” (which in turn must be normatively justified, e.g., through liberties) in an individual case because I need to know, whether the normatively determined requirements of this norm are actually met.

(4.9) Finally, you cannot suspend discourse rational liberalism through an endless discourse of doubt, i.e., a person wants to be bound by the principles only after having discussed the validity of the concept time and again. First, even who does not accept the validity of discourse rational principles is bound to respect etc., since de facto acceptance is irrelevant for the justification of norms. The principles’ binding nature is instead based on a reconstruction of the conditions and implications of human speech (category “reasoning”). In particular, my discourse theory is not a theory of de facto consensual agreement (as many reproach Habermas) but a theory of acceptability in a sense of independence from special perspectives. Second, the required solution of conflicts among humans, which is the function of norms, would be repealed if someone were given a veto power against this approach of justice for turning stubborn (the same problem as with the idea of factual consent). Eventually, every approach of justice would fail. Third, this fictitious objection misinterprets the fallibility of theories: of course, it must always be reconsidered whether a discourse theory indicates the right basic order, and our decisions must remain open for further developments. This does not mean, however, that in the meantime, those provisional findings could be disregarded. Otherwise, norms could never resolve conflicts.

(4.10) One very last point need to be discussed: Is it not true that freedom and autonomy make people unhappy? Is it not true that many people rave about the “feeling of togetherness” (“Gemeinschaftsgefühl”) in closed societies (such as the GDR), the “Golden Youth” in the Third Reich or the “beautiful camaraderie in the Wehrmacht”? Is it not true that many people worldwide feel overwhelmed by the requirement to be “autonomous”? Factually, this might indeed be true. Not by chance did right and left totalitarian regimes of the 20th Century – as anti-liberalisms – arise especially in the historical moment in which the rationalization of our existence carried away more and more traditional certainties. Similar de facto causal relations are likely to explain (in parts) the emergence of today’s anti-liberal terrorism in some countries. There are at least four reasons, however, (two rather formal and two more substantive reasons) why this does not refute a liberal universalism. First, to conclude the validity of a basic order from the actual happiness of the citizens – even if in fact “many people” should be happier in authoritarian states (which I doubt) – encounters the is-ought problem, and as such is untenable. Second, the claim that the happiness of citizens is the measure of political justice, is an unfounded hypothesis, which thus turns out to be arbitrary and a dogmatic setting. Or it ends up in an infinite regress, if someone asks, why the state should give a kind of happiness guarantee to its citizens (and not only the possibility to pursue one’s own way of life). This leads directly to the next point: Third, the happiness of some people in authoritarian states, if it actually existed, is usually bought dearly by the suffering of countless others. Here I think of, for example, the countless people in the GDR who were mentally deformed by fear and intimidation, who were persecuted and had to pay the price in order to allow others to have a “great time at the Free German Youth (FDJ)”. Fourth, the “objection of happiness” misconceives the basic intention of liberalism. For the very idea of freedom is intended to allow everyone of us our own pursuit of happiness. If I really want to, I may also use it to “subordinate” myself to another – such as my partner – or a collective. I do not have such freedom of choice in an authoritarian state. Moreover, liberal statehood is rather supportive of than detrimental to happiness in other aspects, too: For liberal principles promote economic and technical innovations – which can improve the situation of all people (e.g., in modern medicine), if one can overcome their prevailing lack of intertemporal and global justice. Liberalism is an essential condition of our prosperity – it was legal certainty based on human rights that paved the way to capitalism[33] – which also tends to support happiness, at least up to a certain level of wealth. Ultimately, a liberal-democratic organization of our social life is also in a way inevitable. In the West, this is already true today and in view of the rationalization trends, soon it will probably be true worldwide: For once we have learned to decide autonomously about our lives, this is something we can hardly decide to forget. Therefore, there is no way back into the coziness of pre-modern societies – even if fascism and Stalinism were attempts in some way, to undertake precisely that step back to a time before a liberal-individualistic society. Moreover, such coziness was and is usually garnished with economic and technically less mature states, such as a quasi non-existent health care system, government tyranny, and social control to the greatest extent possible – a fact that critics of liberalism irritatingly often ignore.[34]

It is finally worth mentioning that universalism does not require Apel’s and Habermas’s much-criticized assumption that a characteristic of human language was to search for consensus (which is why they impose a discourse obligation on everyone) since my approach refers to “reasoning” rather than “language”. The nature of language “in itself” is irrelevant to me. Only the character of “reasoning” matters. Perhaps we should not call it a “procedural” theory since the unity of discourse and action results in substantive principles.

5. New concept of freedom: The protection of fundamental freedom conditions, multipolarity, intergenerational and global justice

The idea of autonomy and impartiality is thus universally justified as the basic element of social life. On this basis, we could now develop a theory of freedom, democracy, balancing, balance of powers, and political governance (respectively enforcement of justice) that would also enhance some of the aspects discussed.[35] This would also show us that the liberal principles (and their numerous implications in the area of freedom conditions, e.g., welfare, etc.) are not “empty”, but can exclusively and completely determine what is just, including the remaining margins for balancing and rules of procedure. Particularly important is the (new) understanding of freedom. Elsewhere[36], I try to show that the guarantee of freedom must also include a right to elementary conditions for freedom, such as life, health and subsistence. For without them there can be no freedom, so they are (morally and legally) logically part of the concept of freedom. Furthermore, I discuss that, contrary to classical liberal traditions, freedom must equally mean protection against public power and protection from other people by public authorities. Similarly, I attempt to show (morally and legally) that only the basic liberal principles and their numerous implications may be principles of justice. It would be inadmissible for politics to consider additional concerns under the concept of “public interest” (Gemeinwohl) beyond liberties, rights to elementary preconditions of freedom, and the certainly numerous conditions supporting freedom in a broader sense (a spot at a kindergarten, protection of species, etc.). This results from the inescapable principle of autonomy, which forms the idea of self-determination of the individual. The latter may only give way to (democratic, understood as based on a balance of powers) heteronomy, if the self-determination of several individuals is affected. Only these principles follow deductively from the discourse rational basis of argumentation.[37]

Potentiality and other arguments even require a protection of freedom through times, i.e., a protection of young future people by fundamental rights. This leads us to an essential question of modern societies and the real current issue of justice: the extension of the principle of freedom in time and globally. So far, a basic order, which focuses on autonomy, is mostly only intended as an order within every society – even by those authors who think universally. It is not understood as an order between societies or between times. This ignorance leads to serious long-term hazards for freedom – by opening up to question the physical basis of our autonomy. Now there are three arguments in favor of the thesis that morally and legally (as a re-interpretation of the constitutional term “freedom”) we must protect the autonomy of young and future humans – consequently, that there is a “time-neutral” and “global” protection of freedom, including the elementary conditions for freedom, i.e., life, health and subsistence. The possible arguments are linked to today’s discourses, to future discourses and to fictional discourses between different times and derive a “time-spanning” principle of autonomy from all these three levels. Here, I can only briefly introduce the first two arguments for intergenerational justice[38]; they also apply to people living far away in terms of space (global justice).

a)      Potentiality: As we have seen, we owe respect, impartiality, and thus freedom to all potential participants in discourses of justice. In fact, young and even future people are such potential participants. Therefore, they deserve an equal (and equally effective) protection of freedom. A transcendentally inescapable liberal theory of the constitution based on the use of the innate category “reasoning, which is innate to every human as human, is inescapably universal, but also intertemporal (and global). Thus, a universal rationalism is necessarily also intertemporally (and globally) oriented – even though we have not noticed this before.

b)      Eternity: The idea of universal justice, at least since the Enlightenment, is just to make lasting “truths” beyond historical contingencies the basis of all morality. And this is a good thing. For at that point in their life, young and future (and due to the discourse rational approach every human as human) have rights at every point in her life, since she is a being capable of reason, or discourse and autonomy. Note that a right which will definitely occur at some future time, must be significant already today. Because if I damage livelihoods today in ways which prevent young and future people from enjoying their freedom from interference in subsistence, life and health, I will injure them at least at that later date in the future. At that time, however, the damage is irreversible, and thus the affected right could no longer serve its purpose: to guarantee a safe protection against interferences. The same thus founded requirement of protection applies to young people; this leads to a protection against long-term hazards.

This means that future people will enjoy fundamental rights in their future – and that young people enjoy a protection through fundamental rights, which covers their entire life, thus including protection from hazards in the distant future. As indicated, these arguments also build a justification for a global-transnational protection through fundamental rights. Of course, this intergenerational and global freedom is subject to democratic (based on a separation of powers) balancing with the freedom of those people living here and now (and their elementary conditions for freedom as well as conditions supporting freedom of the here and now living). That is why, ultimately, a theory of balancing is the core of a moral and legal theory of fundamental rights. I explain this theory in more detail elsewhere.[39]