The conceptual foundations of decision-making in a democracy

[p. 219]

Part Five
Democratic Argumentation

5.1) Introduction

In a democracy, decisions should be taken as far as possible by consensus. The unsolvable problem of achieving consensus on the norms of justice has been circumvented by limiting our audience to those who give priority to democracy and its principles. The principles of justice are deduced from this objective and form the basis for deducing procedures for taking decisions which qualify as democratic. These procedures, all engaging argumentation, have replaced violence as a means for settling disputes.

Whether or not a proposed decision meets the above criteria of justice is a matter of fact, of truth. Once agreement about the principles of justice has been achieved, all decision-making therefore can be reduced to decisions about facts. Whenever a decision is contested, the bone of contention almost always is a matter of fact, at least in a democracy as defined here. In western democracies, even religious fanatics justify their opposition to abortion or euthanasia either as following from the democratic principles (a fact which engages rules of deductions) or by statements about the expected consequences of tolerating them, also a matter of (expected) facts. Therefore, the way we establish, through argumentation, the facts relevant to a decision is at the core of democracy.

5.2) The Need of Procedures for Establishing Facts.

In the concluding chapter of PART THREE B, p. 123, we have shown that the democratic principle implies the willingness to cooperate towards achieving an agreement about facts which is as objective as possible, for reasons of both justice and effectivity. Unwillingness to do so is incompatible with the democratic principle and justifies the disregard of the consequent arguments or opinions. It was also explained why - given the will - there is a reasonable chance of achieving consensus about those facts which are not dependent for their existence on the perception of human individuals, and why these facts can serve as a common and objective point of reference for all those who are sincerely trying to reach an agreement about these facts which is as objective as we can make it (and thus merits the qualification ‘as true as possible’).

Proof is only needed if no consensus has been achieved and we always must remember that a totally objective proof of the truth of a statement about a fact is impossible. But sometimes we can prove that the truth of a statement is the logical consequence of statements about other facts which have previously been accepted as true. As will be shown in the next chapter, such a proof must be accepted as objective.

[p. 220]

If - as usual - neither proof nor consensus is achieved, we must fall back on procedures for deciding about facts which insure that, in one or more steps, we can arrive at a decision which must be accepted as the most objective achievable under the circumstances. Such procedures must be established in a process meeting the democratic principle and should be established before the need to use them has arisen, as explained when dealing with justice. In cases vital for society, a (qualified) majority will do to establish such a procedure which preferably should be part of the constitution so that we can fall back on it if everything else has failed.

There is no criterion to determine an absolute maximum of objectivity, but we always can, and therefore should, try to improve objectivity by identifying and eliminating unnecessary subjective elements and biases. Doing so requires (a chain of) decision-making aiming to reduce (through logical analysis) contested decisions about facts to decisions about facts on which a higher level of agreement can be achieved, up to a point where the ill-will of dissenters can be proved or - as a last resort - a majority can be achieved which exceeds a preset quorum, hopefully by a large margin. As a last resort, for there is no a priori reason to expect the majority to be more objective than the minority.

5.3) Logic and Mathematics: Shared Conventions to Establish Relations Between Facts.

In any argumentation which appeals to reason, we expect certain rules to be observed which concern the relations between statements of facts: the rules which govern deduction. Some of these relations are allowed or even mandated by these rules, others are forbidden. Most of them pertains to logic and mathematics. All are conventions in the end rest on one or more assumptions which just have to be accepted. They cannot be proved because the very idea of a proof presumes the prior existence of conventions as to what we will accept as proof and what not; attempting to prove these basic assumptions is trying to extricate oneself from a bog by pulling on one's own hair. The basic conventions of logic and mathematics are a set of definitions of:

-	certain ‘words’ (variables: a, x) referring to an item or a class of reality
-	symbols defining relations between variables such as logical constants (and, or, if etc.)
-	values which we accord such statements (in standard logic: true or false).
-	a clearly defined set of operations to create from such sentences (formulae) new, more complex or simpler, sentences.

These conventions are purely formal and would result in a gratuitous game unless the rules thus defined represent relations between facts which we believe to be true. And we do; we accept these conventions because we consider them to be an effective means for dealing with reality. The main property of reality which these rules express is that there is no spontaneous creation or disappearance of objects. The application of logic rests on the conviction that any event must have its roots in other events, an assumption which forms the basis of causality. Logic rejects miracles as a valid explanation of events for which we have no ‘logical’ explanation. It always presumes that ignorance is the cause for the lack of explanation. Logic does not assert that miracles will never occur, but it does state that a miracle can appear in a logical argumentation only

[p. 221]

as a variable whose truth value is assumed at the outset of the argumentation and therefore can never be proved by it.¹

The almost universal acceptance of logic and mathematics in all developed countries is a tribute to the success of their application: their rules have never been shown to be wrong by any event which can be replicated, while everyone can at will create experiments which confirm their correctness. In respect to all other fields of empirical science these rules have the nature of a meta-science. They belong to the field of empirical science in so far as - if we find that they are incompatible with facts - we are willing to revise them: when Euclidean geometry - which is based on the assumption of an infinite universe - clashed with discovery of an anomaly in the perihelion of Mercury and the consequent the theory of relativity, we developed non-Euclidean geometry.

That miracles cannot be reproduced at will renders them unsuitable for decision-making, for in a democracy the only proof of a miracle is the authority of those who have witnessed it. Furthermore, to qualify as a miracle implies that we will never be able to explain the event by the methods of science which we apply to our ‘natural’ (as opposed to supernatural) world. While every individual is entitled to that belief, it cannot be justified as a basis for social decision-making by any method which respects the democratic principle. However unexplainable an event may seem today, we can never justifiably exclude that our failure to explain it in terms of currently accepted ‘natural causes’ may be due to the immaturity of today's science or to a shortfall in the power of observation of all witnesses. In social decision-making, democrats must accept ignorance as the cause of our failure to explain a certain event.

To be accepted as a candidate for regulating democratic decision making, any rule - including logic - must be shown to respect the subjective equality of all members of society. It must be applicable to all arguments of the same kind, irrespective of the interests and opinions of the individuals presenting them. A rule which governs only the relations between interests or between opinions, whatever these may be, cannot make any distinction between the subjects having such interests and opinions and thus meets the democratic principle. Formal languages such as logic and mathematics satisfy that condition for they are valid for all possible statements whose relations they pretend to govern. The facts or opinions expressed by these statements are always represented by a meaningless character (a,x) which can stand for all possible facts, opinions or interests.

As stated, we cannot prove the assumptions on which logic rests nor can we justify the preference of logic to any other set of rules, including anarchy, the rejection of any rule. One argument,

1 As with all fields of human activity, logic and mathematics have many faces; there are various kinds of logic, and there may be differences of opinion about them. Besides the standard logic of propositions and predicates there is for instance Brouwers’ intuitionism; and we have both Euclidian and non-Euclidian geometry. But there are few, if any, legitimate subjects of social decision-making which have properties that place them beyond the acknowledged field of application of standard logic, on the same grounds that Euclidian geometry is perfectly adequate to help us calculate the surface of a piece of flat land on the basis of the description of its circumference. Appealing to the limitations of standard logic when involved in an argumentation about a practical decision is almost never justifiable. Usually this is a symptom of an acute allergy to the decision which would follow from its application.

[p. 222]

however, is as objective as possible because it is based on one, directly observable, fact: the number of people actually accepting the rules of logic and mathematics in argumentation. At a level at which the issue can arise, logic and mathematics are so prevalent as to be practically universal in any argumentation. Muslims, Christians and Marxists all use logic freely in their day to day business, sometimes even in their theological argumentation. This book is full of deductions whose only claim to validity rests on the author's (hopefully) correct application of the rules of logic. It is then reasonable to assume that anyone who participates in the kind of discussion engaged in this book subscribes to these rules, whether he admits it or not. Without such rules of argumentation we could not derive any statements (such as the respect of subjective equality) from any personal choice (such as the choice to live in a society respecting the autonomy of its members). Writing a book like the present one would be a totally gratuitous activity. The whole of western culture would be unimaginable.

Most objections to logic do not really put in cause logic itself. People question the status of logical arguments when their outcome clashes with beliefs and convictions which they do not want to relinquish. Their objection is usually substantiated by establishing ad hoc additional conventions or by referring to an a priori authority, often ‘the’ common sense. Appeal to common sense can be justifiable if it is a simplification of a logical argument on which there already is consensus. But common sense can never provide a justifiable refutation of logic; rather, it is a testimony to the narrow-mindedness or inexperience of its author who has not yet learned that a great many incompatible views can and are defended by appealing to common sense. Religious groups appeal to a ‘higher authority’ whenever they require that any statement in an argument be consistent with a fundamental set of statements such as the Bible or the Koran which must be unconditionally accepted. Democracy must respect their objection to the conclusions of a logical argument if these conclusions clash with their beliefs. But democracy also mandates that this objection cannot be accepted by the whole society (as a rule for argumentation) in its decision-making.

In terms of philosophical rigour the above arguments may not be totally satisfactory. But in view of the vast number of those who do admit the rules of logic, it is justifiable to forego an abstracter, more exhaustive and thus exhausting argumentation and to assume that the reader belongs to that majority.²

While the basic rules of logic are implicitly or explicitly known to most educated people, many are unaware of the fact that in addition to accepting its formal rules, logic assumes that the participants in a logical argumentation adhere to certain rules of conduct, the most important of which is that they do not try to deceive others by making statements about facts which they know to be false. A ubiquitous means for such deception is to first get an implication accepted (‘if A, then B’) and then further on (when everybody will have forgotten

2 Accepting the rules of logic does not mean that we tie ourselves down to a specific set of rules from here to eternity. Today's logic may be improved in a way which maintains the neutrality as to the specific interest and opinions represented by the variables. This issue is mainly of interest to philosophers who are aware that no human construction, whether physical or mental, can be considered eternal.

[p. 223]

that it figured as implicans) assume that A is false. Logic dictates that if A is false, then we cannot deduce anything about B: any conclusion drawn from such an argument is a non-conclusion. Being caught making such a fault is sufficient cause for disqualification. It is a sign of either inability or unwillingness to play by the rules and thus justifies the exclusion of such an argument from democratic decision-making and possibly of its author, if he persists.

The acceptance of the rules of logic or mathematics in democratic argumentation requires qualification. The predominance of formal languages introduces a bias in social decision-making in favour of those who are most proficient in their use. People who are not conversant with the basic assumptions and rules of logic often are aware of their handicap and thus may be reluctant to accept formal languages as the basis of democratic argumentation and decision-making. Democracy mandates that we correct that bias as far as humanly possible, a condition which has been overlooked in today's politics. Society (and thus some individuals) must monitor the content and presentation of arguments to prevent abuses and enable people less at home in abstract arguments to follow the discussions about decisions which will affect them (see 6.3.2, external integration, p. 238) or at least give them confidence that the rules have been observed.

In conclusion, logic by itself can never prove that a certain choice is right; it can only help us clarify what the proposed alternatives really mean and thus protect us against those who by confusing the issue are trying to subvert argumentation to their own interests or who rely on miracles to avoid taking actions which they dislike.

5.4) The rhetoric of Democratic Argumentation: Conviction, Not Persuasion.

In all but the most primitive societies argumentation is the preferred - because least costly - means for obtaining approval of social decisions. The objective of the orator presenting an argument is always to persuade his audience to support of the decision he prefers; he selects his arguments on the basis of their expected contribution to his objective.

Conviction is a specific type of persuasion; it has connotations of rational adhesion, of objectivity, of universality, of intelligence, of... what exactly? The very number and the imprecision of criteria which are used to define conviction as opposed to persuasion indicate that an adequate criterion is yet to be defined.³

3 Kant distinguishes between belief based on persuasion, and belief as ‘holding a statement to be true’ (Fuerwahrhalten) which is based on conviction. Belief is based on conviction if it is grounded in universally valid elements of pure reason. Persuasion is belief that finds its foundation in the subjective nature (Gemüt) of the person holding that belief. (Kant, ‘Die Drei Kritiken’, Raymund Schmit, p.209). It is clear that ‘Fuerwahrhalten’ can deal only with judgements about facts, and more especially about whether a statement or representation of a fact is true in the sense of corresponding to that fact.
Kant's distinction is directly related to the distinction subjective/objective as applied to judgements, and presumes the existence of an objective, if unknown truth. Conviction à la Kant requires that our judgement is based totally on objective grounds, and thus contains only elements which are valid independently of the subject making that judgement. According to current knowledge about the information process, such judgements are impossible, and the notion of an absolute truth must be rejected.
But Kant is correct in asserting that only his methods of pure reason, only methods which we hold to be valid independently of the specific circumstances, can lead us to new information which we must accept as objectively true. Kant therefore only accepts a proof if it is totally based on logic. He excludes from the field of philosophy any argumentation which is not logically compelling, which contains even one element which is subjective and thus cannot be a candidate for universal validity. As rhetoric is mostly addressed to a particular audience, Kant would exclude it as a legitimate means of argumentation. Even if we can never achieve Kant's total universality and objectivity, the requirement in my book of maximal possible objectivity in decision-making about facts comes as close as possible to his intention.
Chaim Perelman proposes another distinction: he calls ‘convincing’ an argumentation INTENDED to obtain the adhesion of every being endowed with reason, and this concept is related to intelligence. Persuasion on the other hand is intended to generate an action and thus is directed at a selected audience, namely those whom the orator wants to prod into that action. Perelman is enough of a realist to add that such a distinction is mainly in the mind of the orator, as nobody can with certitude know what arguments would convince everybody. Only success will tell. The only claim an orator wanting to convince can make is that he has done his utmost best to develop an argument which should convince everybody. (Perelman, ‘ Traité de l'Argumentation’, p. 38)
Perelman says that his distinction will always lack final precision, because the representation which the orator has made of his audience is the result of a judgement which is only human and therefore always subject to revision. That is correct, but he is too modest; for he opposes the imprecision of his distinction to an assumed precision of the distinction between intelligence and will, between rational and irrational. These last two distinctions are just as imprecise as his, for they obtain precision only in regard to a determined and subjective purpose of the individuals involved. While such distinctions may then be of use to the orator, they do not provide others with a method for critically evaluating arguments in terms of their legitimacy, of conviction versus persuasion. My proposal does.

[p. 224]

Indeed, as long as we have no clear view of the nature of information and its role in the life of a human being, nor of the role of argumentation in social decision making, we lack a basis on which to develop such a criterion.

In PART TWO B, about the information process, it has been explained that to maximise our chance of achieving our objectives, our representations must correspond as closely as possible to the reality around us, and therefore be as free from subjective elements as we can make them. Barring errors or misfortune, the action most likely to achieve our purposes is the one which is based on the most objective representation of reality available to us.

The distinction between conviction and persuasion has mostly been approached at the level of the individual's motivations (Kant) and intentions (Perelman). For the subject of this book, it is more fruitful to distinguish between the motivations of the orator and those of his audience: to preserve the autonomy of the members of the audience, the motivations for accepting a decision must be their own.

There are two ways of applying coercion: violence and rhetoric (as the ‘force of argument’, as persuasion). Democracy requires that argumentation about facts aims at conviction. The relevant difference has already been suggested in 1888 by A. Chaignet (La Rethorique et son Histoire, as quoted by Perelman and Olbrechtse-Tyteca, ‘Traité de l'Argumentation, p. 54) ‘Quand

[p. 225]

nous sommes convaincus, nous ne sommes vaincus que par nous mêmes, par nos propres idées. Quant nous sommes persuadés, nous le sommes toujours par autrui’.

Persuasion is the attempt to obtain the adherence of others by any means: by coercion, by an appeal to sentiments and by argumentation; it subjugates the purpose of the audience to that of the orator. The truth of the statements of facts used in the arguments as well as the congruence of the results expected by the orator with the objectives of the individuals composing the audience, are at best secondary considerations. The orator may assure us, possibly sincerely, that he only wants to serve the interests of his audience; but unless we are given the means to check that his arguments meet that intention, we must either accept his word for it and thus bow to his authority, or we must reject his argument if it is presented in a form which prevents us from ascertaining that it respects our authority over our objectives and opinions.

Conviction requires that the decision to accept a proposition of an orator is the same as the one which we would take on our own if aware of all the relevant facts, if the results which can be expected from the decision satisfy our motivations for accepting it. Only then is our consent really voluntary, only then does the orator respect our authority over our own objectives. A prime condition for meeting that requirement is that the representation of reality, the facts leading to the decision, are as objective, as ‘true’, as we can make it. The probability of objectivity and of conviction will be greatly enhanced if the objectives of the orator are, if not identical, at least congruent with those of his audience. Such commonality of purpose is not part of the argumentation itself, but of the circumstances in which the argumentation proceeds, of its social context. Managers who have to obtain the adherence of their staff for a certain project know how important it is to ascertain commonality of objectives right at the beginning of any discussion about the best way to tackle the project.

5.5) The Social Context.

Argumentation takes place in a setting whose most important aspect is the audience to which it is addressed. The main criterion for its democratic pedigree is the relation between the orator and the audience in terms of authority: does the audience expect or accept that the orator has any a priori authority over it? In case of a schoolteacher, the answer may be yes. In social decision-making in a democracy, the answer is no. To preserve the ultimate authority over itself, the audience must have the ability and opportunity to evaluate the arguments of the orator in terms of its own objectives, and its adherence must be obtained through conviction as defined above. This aspect of the social context has mostly been neglected.

In the instrumental view of justice and truth proposed in this book, we would not describe the goal of an argumentation as a ‘victory of ideas’, but as ‘achievement of goals’. In social decision-making, the immediate individual goals should be subjugated to the common goal, which in a democracy is coexistence and cooperation, is to further the legitimate goals of us all and to ensure that all have an equal say in defining these goals.

[p. 226]

In the chapter containing the above quotation from Chaignet, Perelman correctly noticed that the motivation of the orator for choosing a certain argument can rarely provide a basis for deciding upon that argument. As he did not include the nature of society in his social context, he overlooked that it is often possible to prove the one motivation which - in a democracy - is a sufficient cause for disqualification: the intention to thwart democratic decision-making by confusing the issue through introduction of subjective elements, through demagoguery and by distorting or hiding the truth.

Like most authors dealing with rhetoric and argumentation, Perelman approaches the subject only from the point of view of the orator who wants to get his point across. In a normative work about democratic decision-making the appropriate point of view is that of the audience. From that point of view just one condition is important: the argument must, in content and form, be such that the audience can understand and correctly evaluate it in terms of its own objectives.

5.6) Towards Democratic Argumentation.

5.6.1) The obstacles. Summing up the previous paragraph, the distinction between conviction and persuasion highlights that an argument can:

-	improve our understanding of the reality around us and clarify where our interest really lies
-	distort our view of reality and of our own interests to serve the subjective purpose, however laudable, of the orator.

To preserve our authority we must - as stated - be able to evaluate the arguments of the proponent of a certain decision in terms of our own objectives and opinions. Our ability to do so depends on both our current efforts and endowments, and on the properties of the argument. In a chapter dealing with argumentation we can consider our efforts and endowments as given, and concentrate on the ways in which an argument might foil our attempts of evaluation plus what can be done to prevent that.

Rhetoric and argumentation can influence both elements of decision-making: objectives and means. Demagogues can - by appealing to sentiments with a strong affective value or by playing on taboos - disrupt our normal process of critical evaluation and reverse the ‘natural’ priority of our objectives. That strategy is most effective if it can engage our social, integrative tendencies and appeal to morally unimpeachable concepts such as motherhood, anti-racism, freedom, patriotism, emancipation etc. to shield it from critical evaluation. A clever demagogue can easily if unjustly misrepresent any negative evaluation as egoism and its perpetrator as a traitor to the common cause.

He can also mislead the audience as to the relation between means and objective. That is the preferred weapon in the discussions about material issues in today's individualistic and educated western democracies. To decide which policies will best further our objectives is to decide about the truth of facts and the relations between facts. Everybody agrees on the necessity to save our environment and to make the sacrifices which this requires. But translation of this consensus into concrete policies has proven elusive.

[p. 227]

Conflicting or inconclusive studies about the damage done to the environment or about the effects of various proposed remedies enable politicians to hide the subversion of the general interest to that of their constituencies behind a deviating perception of the causes and seriousness of environmental damage and of the presumed effects of various remedies. They can do so because most issues in today's world are so complex, and involve such specialised scientific knowledge, that very few individuals can understand and evaluate them. A politician is usually able to find an accredited scientist to support almost any proposition he might entertain; if not, he often can get away with simply by ignoring statements which are detrimental to his proposals and policies because there is no accredited authority to expose him.

5.6.2) The solution. The purpose of Part Three about truth and its conclusion (p. 123) is to show that it is possible to arrive at a workable and reasonably objective agreement in decision-making about facts provided we are willing to try. Logic is an effective means of safeguarding our control over the arguments, also in political discussions. It requires that we keep objectives and means apart, that we identify the exact nature of the assumptions as well as their role in the argument. To be able to use logic, or other generally accepted formal languages such as mathematics, an argument about facts must be amenable to axiomatisation. In principle all such arguments are axiomatisable.

Clearly not all distortions can be thus eliminated, nor is it necessary to do so. The functional view of information requires only the elimination of subjective elements to the extent that - if not eliminated - they would have induced us to take a different decision. Analysis of the problem situation will show what the decisive elements are and what level of precision is needed to prevent the decision from being misguided by insufficient precision. The effort which is invested in the regulation of an argumentation should be commensurate with the importance of the issue; democracy does not strife for perfection, only for adequacy.

Thus, we do not require that all arguments are always presented in formal languages. On the contrary. As most arguments in social decision-making are addressed to a larger audience, they will have to be presented in common language if the audience is to understand them. Furthermore, if the argument is to be considered and understood by a wide audience, it is not enough to present, in common language, just the facts and deductions which logic requires for drawing conclusions from the premisses. Understanding is more than ‘knowing’ the words. We must arouse the interest required to motivate the audience to make the effort needed to understand our argumentation (see Habermas’ ‘Erkentnis und Interesse’). And we must ensure a sufficient ‘overlap’ among all concerned of the representations generated by our words.

Representations are always constructed by combining already existing information. New impulses simply create new combinations. A new argument will generate (new) information only if it can build on representations which we already have. Whether we like it or not, it will often be necessary to resort to metaphors to mobilise the information already in the minds of the audience towards the construction of new information. The use of a metaphor is therefore totally justified... to the extent that its objective is to arouse the interest of the audience and to promote overlap in the representations made.

[p. 228]

There is no denying that the choice of the metaphors introduces a subjective - and to some extent biasing - element in the argument, as the metaphor by definition always refers to something which is only similar, never equal, to the fact we want to establish. People can, and often do, use metaphors to bias decisions towards their own ends. Because, by using metaphors, we run the risk of introducing a bias even if we have no such intention, this risk simply cannot be eliminated: perfection is not of this world. But we can reduce that risk to a minimum by ensuring that any metaphor in our argument is clearly identifiable as such. And we can and must ensure that the metaphor itself never enters the deduction! That is one of the reasons why democratic argumentation requires that arguments must be amenable to logical analysis. We must choose the ‘best’ possible metaphors, best meaning those which have the lowest affective value and the closest affinity to the facts they intend to convey. While it will usually be difficult, if not impossible, to define what is the best metaphor, we can by counter-examples and logic show that certain metaphors are inadequate or needlessly biasing.

If the argument contains more than just a statement of accepted facts plus deductions prescribed by agreed upon formal languages, we can and must insist that it be possible to assess that the conclusions follow exclusively from such deduction. The presentation of the argument must allow logical analysis by people trained in the subject. If some words have an affective value, we will require that the feelings to which they appeal show no evidence of an intention to bias the judgement towards one of the proposed alternatives.

What if an argument fails to meet these conditions? There is only one sanction which a democracy can legitimately impose on those who reject the principle of democracy and its obligations. That sanction is to exclude them from any claim to the benefits of democracy, specifically from the right to have their argument considered and given the same weight as arguments which do meet the conditions of democratic argumentation. We can ignore the argument, and we can even disqualify its author from further participation in the discussion if we can in a democratic procedure decide that he does not want to play the game of democratic decision making. If he is a first offender, we might give him another chance. But we will have good cause to subject all his subsequent argumentation to extra scrutiny as to its democratic pedigree, and require that he presents them in a form which makes that evaluation as easy as possible.

Note that we only presume that all participants are willing to play by the rules of democratic decision-making. We do not ask that all share our preference for democracy nor that they intend their arguments to convict and not just to persuade. We do however forbid any attempt to thwart our efforts to expose undemocratic elements in all argumentation which is used in social decision-making.

5.6.3) The ethics of democratic argumentation. In the wake of the current fashion of philosophical postmodernism, the attention has shifted from the search for truth and objectivity in science and decision-making to rhetoric and ethics. McCloskey's ‘Sprachethik’ (McCloskey 1985, p 24; 1994, p 990) requires that the conversation be ‘honest’. The rules he proposes are: ‘Don't lie; pay attention; don't sneer; cooperate; don't shout; let other people talk; be open-minded; explain yourself when asked; don't resort to violence or conspiracy in aid of your ideas.’

[p. 229]

All but the first are actually prescriptions for good manners; they can be established and monitored ad hoc by the participants themselves according to the purpose of the dialogue. Methodological, law-like rules are required whenever the argumentation has consequences for third parties, usually because the results are used in decision-making. Only the first of McCloskey's rules (be honest) has any bearing on the really important element in decision-making: the content of a statement. I will use integrity instead of honesty because integrity also has the connotation of ‘without ulterior motive, hidden agenda’.

The inescapable subjectivity of all information and the inevitability of ethics in any statement about social norms are major contentions of this book. They are also the major tenets of postmodernists. Had they stuck to them, they would have made a crucial contribution to modern science and philosophy and been a most welcome ally. Unfortunately they succumbed to the ubiquitous tendency towards hegemony and hubris so characteristic of new developments in philosophy and social science. They had to eliminate or at least subjugate the main competitor of rhetoric: the search for truth and its concomitant prescription of objectivity. Their weapon is the denial of a possible existence of an observer independent ‘reality’ (See Volume Two: ‘the confusion about reality’, p. 423). Analysis of their argumentation shows that it is based on non sequiturs and petitio principiis; it is a perfect example of how to immunise a proposition against all possible criticism and enable its proponents to justify all interpretations of their creed which they consider opportune. By the standards deduced from the choice for democracy their argumentation is profoundly unethical.

As stated, the democratic principle and the rules for argumentation presented, including the quest for objectivity in establishing facts, are ethical concepts. The ethics of democracy must be made explicit and ‘interiorised’ if democracy is to flourish. That must be a cooperative venture.

This part only provides abstract concepts and recommendations. The bridge towards reality is the word ‘we’. Who is this ‘we’? On whose shoulders rests the responsibility for ensuring the democratic pedigree of the argumentation? That is the subject for the concluding chapter of this volume.