The role of ideals in reasoning

One year after I put the first fragment of it on this blog, and two-and-a-half years after the MA thesis it is based on, here is the first full version of my article on "the role of ideals in reasoning" at last. I will be very grateful for any comments and reactions. If you prefer a .doc file (8 pages) you can download it here.

The role of ideals in reasoning

My claim in this article is quite simple. Ideals perform a constitutive role in reasoning. By ‘ideals’ I mean conceptions of how things should be, sometimes hard to articulate and even harder to substantiate. By ‘constitutive’ I mean that without such conceptions we are not really reasoning, but merely acting according to some rules or standards of reasoning – which is not the same. The ‘constitutive role’ is not meant in the sense of being there before we develop a faculty of reasoning, or performing a guiding role in the acquisition of a set of rules. The role they exercise is in the everyday uses of reason, making decisions, giving meaning.

The argument for this claim is as follows. Both in making decisions and giving meaning, we are performing motivated actions. The motivation, however, often leaves something to be desired. We are not able to fully explain why we did something, or what we mean by some expressions. Now this is not a sign of fallacious reasoning, but of the fact that reasons are underdetermined. They must be, because they are something we can disagree with. A law or a formula is not an argument: those you can apply or not. Of course, laws and formulas are immensely powerful tools for reasoning. But when employed as arguments, they merely function as discussion-stoppers, precisely because there is nothing to disagree about, except whether this specific one works and can be applied. Formulas, in short, do not count as motivations.

Motivations take into account two aspects of things: how things are and how things should be. A motivation that does not take into account how things are is plainly ineffective. But a motivation that leaves things as they are motivates nothing. A solid motivation, then, relies on a conception of how things should be, but one that is under constant scrutiny in view of the facts. Without this scrutiny, it declines into the kind of ‘if only’-reasoning that is often denounced as ‘idealism’. In so far as this is a label for defective reasoning, I agree with the diagnosis but disagree with the name.

The problem with ‘how things should be’ is that there is really no way to find out. (‘Things’ here is shorthand for life, reality, or any object intended in an action or expression.) As far as I’m concerned, it is okay to trace the origin of ideals in physical drives, empathic responses, and the spin-off from restless minds. They are certainly not ‘things’ in the sense of isolable objects; identifiable best, perhaps, by the words in which they are expressed. There is, then, no such thing as a ‘science of ideals’ but at best an analysis of how the terms that denote ideals are used and abused. Such an analysis will always be a scrutiny and often an unmasking; it comes closest to a ‘critique of ideology’ or the original meaning of idéologie.

Moreover, a conception of how things should be prompts the question of why things aren’t that way already. If sufficiently clear and coherent to be formulated, conceptions of how things should be will often demand something impossible. But this, again, is not a defect. Soyez réaliste, démandez l’impossible – the slogan of the soixante-huitards sums it up neatly. A proper motivation demands something that comes not of itself. ‘How things are’ can be a reason not to do something, but not a sufficient reason to do something. One way of identifying the conceptions underlying these motivations as ideals is precisely that they are supererogatory. In that sense, every ideal demands too much.

Shaping yourself as reasoner

In this article, I will not be concerned with ‘proving’ that there is a constitutive role for ideals in reasoning through a set of examples. The issue is rather how to conceive of reasoning. This, my claim goes, is itself an issue of how things should be as much as an issue of how things are. ‘Proving’ the role of ideals in reasoning would itself be a somewhat despotic prescription on how to think. The aim is not to develop a ‘fuzzy logic’ contaminated with ideals, or to exclude reasoning from formal description altogether. Logic, formal semantics and linguistic theory are all perfectly respectable pursuits for which ideals provide no substitute.

But what does it mean to be a reasoner? Unsurprisingly, I argue that it is not just a matter of following a set of rules, consciously or mechanically. Rather, it is a creative and a tragic process of shaping yourself, in the course of communicating, making decisions, accumulating and testing beliefs, wavering between self-invention and self-correction. The process is creative in that it sometimes brings you to do or say things that you could not predict or anticipate; it is tragic not merely as a struggle with your irrational inclinations, but also with the intransparency of your own cognition and decision-making, and the limits of effability.

The tragic and the creative aspect of reasoning come together. What happens is, we see ourselves act, we hear ourselves speak. In a way, this is like prodding your memory or wondering whether you saw or understood something correctly: adopting a third-person perspective to your own mental mechanisms. Now the mechanism may be as mechanical as they can be, but the mechanisms are not the reasoning process; they are what makes it possible for you to reason. The standard argument in philosophy of language is that reasoning is normative and public; but that falls short of capturing the creative element of reasoning. Public norms are something to take issue with rather than to rely on. When it is innovative, as well as when it operates on conceptions of how things should be, three other aspects of reasoning come into play: that it is goal-oriented, self-corrective, and able to cope with novel circumstances. The goals themselves may be ‘wired in’, even as impulses. What matters is that, as a self-correcting mechanism that can make conceptual innovations to cope with novel situations, the mind does not stop at these goals, does not hit rock bottom. You may not know what to say or do – that is what comes with seeing yourself act and hearing yourself speak – and rightly so, for complete lack of doubt is a reasoning defect. (This is why there is no such thing as ‘uncritical knowledge’, and why set and formalized ideals, in the absence of further reflection upon ends, cease to perform a constitutive role and become ideology, another reasoning defect.)

Now this doubt indicates a discrepancy between how things are and how things should be. And such a situation, reasonably, calls for innovation in order not to remain stuck. You can remain stuck all the same, for lack of innovativity, because you keep on dreaming, or because there is no clue to be found – all that is tragic in its own way. But ‘tragic’ need not apply only to far-fetched existential scenarios: in a sense, every choice is tragic, because if there were knockdown reasons for one option it would not be a proper choice. It is tragic that you will not make a better choice in a restaurant by studying the menu longer, or that you will spend more time, statistically, in the wrong queue. It is tragic that you will find the mot juste after the opportunity has passed, late at night in bed preferably. Such minor tragedies are part of shaping yourself as a reasoner. Finding a way of coping with them is a way of innovating upon yourself, hopefully for the better. Here the petty bourgeois and the bohemian are on a par, even if the latter’s conceptions of how things should be are more sweeping.

Giving meaning

There seems to be something missing here. To spot a discrepancy between how things are and how things should be is not yet to have a developed conception of how things should be. And what is understood in this article as ‘reasoning’, anyhow? As outlined in the previous section, I understand reasoning as something that is involved in cognition and decision-making, and that is self-corrective and creative. Furthermore, it was stated that reasoning should not be confused with the mechanism that makes reasoning possible; therefore, it is hard to define it as a measurable “object of study”. Rationality is a quality; it is not even very controversial to call it an ideal.

Now it is important not to confuse the fact that people are not optimally rational – whatever ‘optimal’ rationality would be – with a proof of the role of ideals in reasoning. For although it helps to assume that people know what they are saying and doing – what Donald Davidson has called the ‘principle of charity’ and at some point even ‘the constitutive ideal of rationality’ – we also know that they lie, err, joke, talk nonsense, chat, have nothing better to do or say, tinker, and strike a pose. Man merkt allmählich leicht, daß auch kluge Leute bisweilen faseln! They may be driven by coherent motives, and there is certainly some idealization in the interpretation of what they say and do, but that does not mean that they are driven by a desire for the true and the good.

Rationality is just one of the concepts that we use that are idealizations. The same goes for number, measure, truth, meaning, as well as right, good, beautiful, and maybe even cause and consequence. One could even claim that every concept that is not ostensible is an idealization – and then, what is ‘ostensibility’? But that does not show that there is a coherent conception of how things should be behind these concepts – in fact, it only leads into regress.

Still, there are some idealizations that have attributed traditionally with a privileged role in explaining how concepts have meaning. These are, indeed, coherence, optimality, and normativity. They are idealizations in that there is no conceivable way to measure the overall coherence of one’s ideas, the extent to which an expression is le mot juste, and how reasoning is governed by norms. That people’s ideas are inferentially coherent ­– that is, if you know one thing you know a lot of things – is an established fact; but it is not as if knowing one thing is knowing everything. The overall coherence is something to be desired and pursued rather than actually acquired. There have been attempts in pragmatics to explain how people can communicate without making everything explicit in terms of optimal relevance; as a formal description of inferential mechanisms, these attempts have failed. There are other considerations, such as prosody, memory retrieval, game theory, and sheer enthusiasm that play a role in making yourself explicit. Still, both Grice’s maxims and the “principle of relevance” into which Sperber and Wilson (1987) collapse them have some intuitive appeal: yes we do want to make sense. Normativity is equally elusive: of course we say (and do) a lot of things simply because that is how you do them – but that does not create some isolable entities called norms, and it doesn’t mean you can’t flout them or opt out. One might attribute a constitutive role to norms in reasoning in terms of background assumptions and habits; but they only acquire force when there are reasons for them.

Now there is a way out of the regress of calling everything an idealization: and this is to weaken the explanatory claims that have traditionally been heaped upon coherence, relevance, and normativity and to describe their role in terms of shaping yourself as a reasoner instead. In that sense, giving meaning is shaping yourself. You do so by motivating why you do things, finding the best way of expressing, and checking the coherence of your ideas by submitting them to critical scrutiny – including your ideas of how things should be. This conception of shaping yourself as a reasoner has some affinity with what Rorty (1979) and McDowell (1994) describe as “Bildung” – except that they describe it as a process of acculturation, and I rather view at as an ideal, which is rather hollow if not subjected to critical scrutiny.

Conversely, one might even claim that reasoning itself is giving meaning rather than performing such-and-such operations: one the one hand, there is no universal “reasoning” operation, and the rules of logic are rather unfit for explaining cognition and decision making; on the other hand, for all the problems of making explicit what you were after the act of reasoning requires some awareness of what you’re doing. Now I don’t claim that awareness is the defining property of meanings (meanings may very well be “not in the head” – at the prize of a considerable idealization) but it is a crucial element in the critical scrutiny of your ideas. Maybe we make decision before we know what we do or talk automatically – a great deal of language processing indeed goes unreflected – but if it is beyond scrutiny it is meaningless. The things that are uncritically rote learned and drilled in are meaningful only because they enable you, at a certain point, to think and act autonomously.

Doing things with ideals

If ideals are to perform a constitutive role in reasoning, how are they to be developed, made explicit, and last but not least, acted out? Fernando Pessoa wrote that “who seeks ideals does not have them”. That is indeed a paradox we have to live with, equipped with a mind that entertains motivations and pursues coherence, but that can also opt out and say “so what?”. The same awareness of contingency (that things could be different) and critical attitude (that they should be different) that lead us to develop conceptions of how things should be also lead us to abandon them. Left at that, the role of ideals in reasoning is that of a rest-product rather than a constitutive force.

However, it is not as if we first have raw impulses and desires and then concoct ideals from them by fitting them to the facts. To view them as conceptually empty would mean, in extenso, to deprive them of all relation to the world. (Meanings may not be in the head, but desires and impulses certainly are.) Rather, motivations already come content-laden. And conversely, without a sense of relevance there would be nothing to focus on, nothing to notice in particular: perception would be lost in raw data. It takes some sorting out to arrive at distinct conceptions of how things are and how things should be. Only, the sorting out is considerably easier with regard to how things are because facts are typically supported by the world whereas ideals typically aren’t.

The issue is not to develop a maximally coherent and maximally imperative conception of how things should be, and stick to it: that would rather be an invitation to cognitive dissonance. The issue is rather to develop a sense of what you’re up to. And that may be a fact of life, but with regard to the question how to conceive of reasoning, it also comes as an imperative. It is mandatory to conceive of reasoning as a self-corrective and creative pursuit, even if real-life reasoners do not always live up to the demand. From the perspective of practical reasoning, demanding that every action be informed by an ideal is grossly supererogative. But to be self-corrective and creative isn’t. One could shrink the role of ideals in reasoning to a very simple demand: do as you already do, only better.

Is that all? No. Reasoning is not only about individual choices and actions. There are much more complicated reasoning processes going on in dealing with other people – typically, trying to make sense of what they are up to, which involves a further scrutiny of your own motivations. And when it comes to decisions that affect other people, it is mandatory not only to have an eye for everyone’s interest but also to have an idea of what you’re trying to achieve – that is, good politics involves a debate on how things should be. Last but not least, to claim a role for ideals in reasoning affects one’s conception of philosophy – a particular sub-genre of reasoning that does not fare well in producing facts or decisions but that seems best suited to the systematic pursuit of how things should be.

The role of ideals in reasoning is thus different within different domains or styles of reasoning. On the level of individual decision-making, it comes down to a process of ‘shaping yourself as a reasoner’ that has been described above. However, this process largely takes place on the social level – since most of reasoning and decision-making is with regard to others. Here, it is not merely moral but also strategic considerations that come into play. And though moral considerations are crucial to social interaction – there is no way of completely sidestepping them short of outward war of all against all – they will always demand too much in the face of strategic considerations. The phenomenon of “paying lip service” is an excellent illustration of the paradox we have to live with, in that it implies both an acknowledgement and a denial of the role of ideals. On the level of political deliberation, these considerations are muddled further. On the one hand, conceptions of how things should be are much more explicit here than they are on an individual or social level, and divisions between political parties are often along these lines; but by the same token, ideals become a matter of rhetorics, ideology, and strategy. The debate will seldom convince anyone, and virtually never lead to new ideas. Still, there is no politics without debate, and the debate is not as much about how things are as about what should be done.

It is not merely strategic considerations that make public reasoning about ideals problematic. There is another complication in the further paradox that lies in pluralism about ideals. Consistent reasoning from conceptions how things should be implies that things-should-be-so and hence I-will-do-this. But that is not consistent with the ideals of others. Consistent social reasoning implies negotiation, suspending judgement, and in the best of possible worlds, public deliberation. But each of these, when it comes to being true to your ideals, are forms of functional hypocrisy. The solution, though qualitate qua an inadequate one, is to make social reasoning part of your reasoning about how things should be: to pursue open discussion, learning from each other, and effective communication of opinions as an ideal in itself. For all that social interaction does to shape and enrich your view on life, this way of reasoning still is a give-and-take that calls for a different kind of creativity, i.e. creative bookkeeping. No amount of openness to others will take this paradox away.

Openness is a virtue, but it is not enough. It is a crucial element to reasoning on all levels, individual, social, and political (in the sense of openness to new experiences and impressions, openness to others, and democratic openness, respectively), but it does not give reasons for action or judgement. “The more, the better” is not a maxim when it comes to giving reasons; to identify a good and sufficient reason you have to apply Ockham’s razor. If you one knock-down argument you don’t need the rest. Openness only leads to some new and positive conviction when it is combined with another virtue: the art of selection and judgement.

Here, in the creation of new ideals and the art of selection and judgement, there is a role for another domain or style of reasoning, namely, a role for philosophy. And that is what the closing section of this article is about.

A role for philosophy

In this article, I argued that ideals can be informed by all kinds of knowledge about the world, but cannot claim the status of facts. The same, analogously, applies to philosophy. Even stronger, I would claim that for both, an openness to all kinds of knowledge about the world is mandatory. Your insights must come from somewhere. You cannot make a claim about how things should be, or a philosophical claim, simply by closing your eyes and thinking very deeply. The claim must be argued for, it must be made plausible. An appeal to authority or inspiration is a particularly weak argument here. Philosophy is not an artwork to be admired. Still, no amount of facts will prove it. In this regard, philosophy is somewhere in between art and science, a sort of ‘art of thinking’. (It is unlike art, however, in that a philosophical claim can always be argued about, and with a piece of art this need not be the case – if it were a claim to be argued about, one might better just as well state it instead of paint or dance it.)

The analogy between ideals and philosophy is not complete. There are ideals that are not particularly philosophical – political, religious, or artistic ideals, or more profanely, the ideal sports car, or the ideal cake. Conversely, not all philosophy is strictly about ideals – even though it arguably is about how things should be, how to conceive of things, what to do, and such normative issues. With regard to the role of ideals in reasoning, I would argue for a wider conception of philosophy: not an academic discipline based on ‘conceptual analysis’ but the more systematic and sustained pursuit of conceptions of how things should be.

The good news is that we already do it anyway, only less systematically. There is a modest role for philosophy in informing “informed opinion”, spelling out the logical and moral consequences of this or that position. But this is a necessarily modest role in that it only comments upon ongoing debates. There is a less modest role for philosophers in actually starting these debates, and giving the good example in having more than just another opinion. This is a pragmatic conception of philosophy as more or less identical with being a ‘critical intellectual’, someone who feels obliged to know what he or she is talking about and looks beyond vested interests. And there is a more romantic role for philosophers as innovative thinkers, whose basic motive is to know everything, try all and retain the good. This is an impossible task indeed, but it is a reasonable demand if you want to have something new to say on how things should be – soyez réaliste, demandez l’impossible indeed.

None of these roles is strictly delineated from what all thinking people do. In that sense, the bad news is that it leaves nothing identifiable as philosophy, except a certain foolhardiness. As Wittgenstein said, when we are talking philosophy, we are like wild men screaming and shouting – and indeed we should. The only thing that distinguishes philosophy in this sense from ‘having an informed opinion’ is the willingness to wax philosophical, to break the rules of conversation and go beyond the matters at hand. The only thing, on the other hand, that distinguishes philosophy from mere speculation or even chatter is the ongoing commitment to explain and revise your opinion, and not to put it beyond scrutiny. It is not specifically the task of philosophers to show that everything is more complicated than it seems. Experts can do that. Having an opinion is always a wager no matter how much you know, and still it is better to have one than not – provided it’s open for correction.

A closing note

Why should there always be new ideals? There are good reasons not to revise one’s ideals: because they are inconclusive anyhow, it might not be in your interest to do so, and for the sake of moral clarity. There are counterarguments against all three, but they are, indeed, inconclusive.

We do it anyhow. Reasoning works by guesswork and self-correction. That is an argument to renew your ideals even though they are inconclusive anyhow, but not when you have an interest not to.

Not to renew your ideals is self-defeating. This is true of the strategy adopted consciously by political brokers, unconsciously by religious zealots, and commodiously by most people in smaller matters, simply to stick by what they think is good. The world is changing, and new information calls for your attention daily; if this does not make you change your views you are rendering both irrelevant, the new information because it does not evoke a reaction that affects your motivations, and your views because they lose correspondence to what goes on in the world. That said, there are good reasons to go into denial or simply choose not to care, and not revising your point of view is part of how the mind works as well. Some degree of conservatism is needed to keep your opinions coherent and consistent, if only for the sake of moral clarity.

But clarity is not enough. Your ideals must make a difference. There must be a motivation behind your opinion that makes it worth hearing, that adds something to what has already been said. It takes creativity to make yourself explicit in a relevant way. On the other hand, the truth need not be interesting, and if there are moral truths they are fairly plain: don’t be cruel, poverty is bad, live and let live. But if these are moral truths, they are also mum and apple pie.

This essay started from a critique of a formal conception of reasoning and ended with a set of normative judgements about the art of thinking. In a way, this has been an attempt to give a new meaning to an old moral – sapere aude! – and an outdated ideal – “Bildung”. The sections on how to do things with ideals and a role for philosophy give some indication of how to proceed. How to elaborate the role of ideals in reasoning in greater logical and psychological detail, however, is a question to which I have no answer. I hope that this article provokes some thoughts.