How Can the Field of Value Be Made More Pragmatic?

Value as Ways of Behaving

Nathalie Heinich rightly proposes escaping essentialism with regard to values; she attempts to do so by considering values as individual and collective mental representations, and by bringing sociological enquiry to bear mainly on the acts and operations of evaluation and justification. It is necessary, however, to first better understand the act from which this recurrent essentialism proceeds. Basically, it consists of transforming into a substance something which in fact has only a functional or modal reality. Anyone who makes value a specific category of existing things, designated by abstract names, commits a sophism of this type. With regard to mind, matter and conscience, Dewey wrote in 1925:

It is a plausible prediction that if there were an interdict placed for a generation upon the use of mind, matter, consciousness as nouns, and we were obliged to employ adjectives and adverbs, conscious and consciously, mental and mentally, material and physically, we should find many of our problems much simplified. (Dewey, 2012 [1925]: 97)

Two decades later, he made the same methodological recommendation with regard to value, a term he advised using as an adjective, to avoid falling into the traps laid by the use of abstract nouns (Dewey, 1949). In short, he believed that abstract nouns merely recapitulate or summarise the qualities of the experience.

Ordinarily, an adjective qualifies a noun. But an adverb is also an adjective, or more exactly an adjective that qualifies a verb: it specifies the course of action indicated by the verb. Adopting an adverbial approach to values involves first considering them as situated courses of action, implying ante-predicative qualifications that are observable in the attitudes and behaviours an organism adopts in its transactions with its environment. In these transactions, it constantly puts a value on objects, events, situations, behaviours, practices, relationships, etc., including the sequence of its own accomplishments, along with their finalities and resources, results and consequences. These ‘valuations’ are functional, in that they contribute to guiding the changes that constitute behaviour.

It is probably worth specifying three points connected with this adverbial approach. First, that an environment is not merely a set of resources and constraints (as is often thought in sociology) but also a ‘co-operator’ – the operations through which our behaviours and our activities take shape are distributed between the body and its environment (as designated by the term ‘transaction’). Second, that behaviour or conduct is determined by guidance and control over changes in an activity that is a sequential, serial process – changes that are more or less coordinated with those that occur concomitantly in the environment (Dewey, 1930). Third, that the qualifications made in a judgement transform qualitative totalities that are directly felt in the immediate experience into objects of thought (Dewey, 1931).

It is a fact that ‘everything under the sun’ can be ‘valued’, i.e. qualified in terms of value. But qualifications in terms of value go far beyond actual evaluations and value judgements, and therefore do not necessarily create ‘object-values’ or ‘goods’ (when I ‘value’ the means for my actions, I do not make them ‘object-values’). Qualifications in terms of value are largely purely practical, behavioural and situational; they nevertheless have a face or physiognomy, which makes it possible to identify them; they also arouse expressions: that which we are inclined to say or are prepared to do in their wake. We have a whole series of verbs to describe this type of qualification: set a price or value on something, give it importance, cherish, esteem, prize, appreciate, admire, honour, approve, revere, give weight to, defend, be devoted to, take an interest in, concern oneself with, and so on. All these verbs are positive, which reflects the positive connotation of the verb and noun ‘value’. We must however also take into account the bivalence of qualifications in terms of value: verbs such as disdain, decry, deprecate, detest, undermine, hate, despise, loathe, suspect, etc. all express negative qualifications.

The words ‘valuer’ and ‘valuation’ do not exist in French (it is not possible to use the word ‘valorisation’, which has another meaning), but I feel it is necessary to use these Anglicisms [in this originally French text] to take account of the fact that actual valuations and ‘value judgements’ only represent one of the ways of producing qualifications in terms of value, and also of the fact that the behaviours attesting ‘valuations’ have a biological source and a motor-affective dimension. Since Nathalie Heinich refuses to consider values as natural occurrences, I shall discuss her point of view. Indeed, although she does note the plurality of qualification methods, among which she includes ‘attachments’, she nevertheless favours the evaluative acts and evaluative operations that underpin judgements – which, to my mind, over-reduces the field of value, at the risk of introducing a cognitivist bias into the analysis (a denial of the anchoring of cognition – thought, reflection, judgement, etc. – in an immediate thoughtless experience).

And it is not only values and qualifications methods that are plural: value systems are also plural. There are scientific, religious, philosophical, moral and aesthetic types of ‘valuation’. If we take the case of the latter type, it must be admitted that it is not exclusive to the field of art. An ordinary activity may have aesthetic value, as may a discussion or conversation (see the example of the recruitment interview Dewey gives in Chapter 3 of his Art as Experience), or give rise to ‘valuations’ of an aesthetic nature. When I use a hammer to hammer nails into a plank of wood, I am able to appreciate a number of things in my experience: the way my hand suits the shape and size of the shaft in a comforting fashion, the corresponding metal mass, its total suitability for the task at hand (it could have been too light or too heavy; the material it was made of could have been unsuitable for the task at hand; for instance, if it had not been made of steel but of wood or rubber, something would have disturbed me and prevented me from appreciating it), and so forth. The experience itself is ‘valued’ for its development, in the same way as is the activity for which I use my hammer: positively, if everything goes without a hitch, if the sequence of operations is fluid and the result satisfactory – the ‘valuation’ is then of an aesthetic nature; negatively, if all goes awry, if the activity is intermittent, disorganised or scattered, or if the result is disappointing.

Ludwig Wittgenstein (quoted in Chauviré, 2004: 53) referred to such spontaneous ‘valuations’ as ‘aesthetic reactions’: ‘perhaps the most important thing in connection with aesthetics is what may be called aesthetic reactions’. An example of an aesthetic reaction is non-satisfaction or discomfort when faced with an object – the drawing of a door, for example: it may be too high, too low, badly proportioned in relation to the house-front or to the windows. When I react in this way to the drawing, it is an immediate motor-affective ‘valuation’. I may, however, formulate the reaction – by saying, for example, ‘it’s too high’. I then produce a critical evaluation of my drawing at the same time as I express my dissatisfaction. I could also express it by means of a particular facial expression or a gesture. While expression is not necessary to feel a lack of satisfaction or discomfort, nevertheless when I am not satisfied by something – my drawing of the door for example – I am inclined to say or do such and such a thing. For Wittgenstein, aesthetic reactions are an extension of natural reactions, i.e. reactions that do not proceed from thought, calculation or interpretation, but belong to ‘the general facts of nature’ (without reacting ‘naturally’ to rhythm in a certain way, we would not have the music and dances that we have). This is a pre-linguistic mode of behaviour, but it may be formulated and refined using language.

Unlike evaluations, ‘valuations’ are not a specific type of act and they are not based on operations. They form part of the normal course of behaviour and are incorporated, as means of guidance and control, in what we do and undergo. They are purely practical and manifest themselves firstly in modes of behaviour. Consequently, there is a risk that a sociology that focuses more on evaluative acts and what are called value judgements will miss a large part of ‘valuations’, particularly because it also often forgets that value judgements, when they are authentic judgements, i.e. when they proceed from a considered examination and are not merely immediate reactions or the expression of first impressions, are above all practical judgements: they serve to direct the changes that constitute the behaviour, by the formation of desires and by the anticipation of probable consequences, and they inform subsequent ‘valuations’.

This last point is important: value judgements influence, and sometimes even change, immediate ‘valuations’. I therefore do not share the point of view of Nathalie Heinich (2017b: 312) that ‘the sociology of values . . . has nothing to do with any kind of ethic’ (la sociologie des valeurs . . . n’a rien à voir avec une quelconque éthique). Such a position can only be maintained by forgetting that value judgements are practical judgements, i.e. that they concern the organisation of behaviour, and particularly the determination of the ends to be pursued – they give activities their direction and check the process of their accomplishment – and that often they redirect immediate ‘valuations’. Under this scheme, the values resulting from these evaluations are by turn ends and means. They are the means for apprehending the consequences, and therefore for defining new ends. In general, the ends and the means are themselves ‘valued’.

On the Refusal of Naturalism

Nathalie Heinich explains to us that values are ‘facts of culture and not of nature’ (faits de culture et non pas de nature). I would rather say that they are both at the same time, and that there is no justification for refusing a degree of naturalism. Heinich’s position is understandable, since naturalism is often reductionist – for example, when it considers the brain and hormones to be the foundation for social life. With regard to ‘valuations’, however, it cannot be denied that they have a natural biological basis, i.e. that they are part of the instinctive behaviour of living beings in their interactions with their environment. To ensure their survival, in fact, they constantly carry out selections and rejections, on the basis of attraction and repulsion; they move towards what attracts them and seek to obtain it, while they avoid or even eliminate what does not suit them or what they dislike; thus they adopt behaviours or attitudes that demonstrate that they ‘value’ the elements in their environment positively or negatively. These behaviours include, at least for evolved animals, taking care of offspring, and feeding and protecting them, because they endow them with value, and even behaving empathetically towards their fellows (as suggested in some recent studies in ethology).

As this description emphasises, these instinctive behaviours have an affective dimension (liking/not liking, appreciating/hating, etc.) and a motor dimension (move towards, promote, approach; move away from, prevent, distance oneself from, eliminate, defend oneself). This motor dimension is important: it corresponds to both the effectuation of movements and the dependence of feelings on motor activities. When we enjoy something – eating a meal, for example – our enjoyment is based on a sensory-motor interaction with the object that is being appreciated, an interaction that is a combination of activities (what is happening in our mouth, on our tongue, on our palate, etc.) and passivities (what we undergo).

However, purely instinctive behaviours lack the intellectual element that would confer on them a dimension of desire. Desire appears when the living being includes its affective-motor ‘valuations’ among the purposes of their actions. As in John Dewey’s example: a hen that would consciously aim for, as the purpose or result of her activity, the hatching of the eggs she broods would ‘value’ both the eggs she takes care of and the activity of taking care of them. She would bear a desire. The affective-motor ‘valuation’ would then become a ‘valuation’ that is at one and the same time ‘affective, ideational and motor’ (Dewey, 2011 [1939]: 115).

Affective-motor ‘valuations’ are not absent among humans, but they are made more complex when such an intellectual element is incorporated and developed; incorporation enables them to be changed not only into ‘desires’ but also into actual evaluations, including evaluations of themselves, at least for all practical purposes (for example, do they deserve to be maintained as foundations for action?). An immediate ‘valuation’ may thus on reflection prove undesirable or problematic; reflexive evaluation transforms it in the same way that it transforms the object that is ‘valued’. There is thus both a continuity and a discontinuity between immediate affective-motor ‘valuations’ and the actual evaluations, and they include value judgements: ‘Judgment of values, in short, is the deliberate development of an aspectual constituent of the more direct prizings and cherishings that human beings as living creatures must and do continually engage in, and under such conditions that at first they are relatively “thoughtless”’ (Dewey, 1949: 354).

This reflexive development is, however, merely one of the possible forms evaluation may take. It may also be nothing more than the expression of a first impression; it would then lack the solidity it has when it is the conclusion of a deliberate examination, based on a thorough perception. At any rate, all evaluations, including value judgements, have as their starting point impressions of a qualitative nature that are produced in us by things and events (which are in the first instance ‘felt’). We should note that a properly conducted evaluation, going beyond an immediate reaction, does not necessarily seek to attribute a value to an object: it is not so much concerned with setting a value on the object in question as with achieving a more thorough perception, by grasping its objective properties; this does not necessarily lead to a judgement on the value of the object as a whole, although it may do so.

We may therefore naturalise the field of value without difficulty. This naturalisation would nevertheless involve a cultural naturalism, since humans are acculturated beings. Their ‘axiological experience’, like any experience, is a matter of immediate interactions with an environment, interactions in which what is undergone and what is done are combined and balanced. These interactions are governed by impulses and habits. It is this intervention of habits that channels impulses, and as a result the background to the immediate affective-motor ‘valuations’ is largely social and cultural. The acquisition of habits does in fact include the formation of sensory-motor coordinations as much as it forms emotional and intellectual attitudes; it shapes our instinctive affectivity just as much as our many and varied reactions to our surroundings or to the things that happen to us. These habits are formed in a socio-historical milieu; they are inculcated from a very young age by the behaviours adopted by other people in the same social environment – approval/disapproval, reward/punishment, injunction/interdict, etc. In this way mores, customs, institutions and prejudices condition (causally, we may say) the immediate affective-motor ‘valuations’, not to mention all kinds of possible manipulations on the part of the holders of economic, political or cultural power. Analysis of what conditions the ‘valuations’ and evaluations in this way and analysis of their consequences would be tasks for the sociology of values (supposing that such an academic specialisation existed).

Nathalie Heinich refers to ‘axiological experience’. If by that she means a specific type of experience, I believe she has taken a wrong direction: what I have just explained with regard to the omnipresence of ‘valuations’ in behaviours and activities would tend to prove that the ‘axiological’ dimension is more of an inevitable component of any experience than a specific type of experience.

Emotions as Sensors of Values

The emotional component of ‘valuations’ (I have called them ‘affective-motor’ or ‘affective-ideational-motor’) and evaluations is essential. Emotions inevitably come into play in value qualifications, essentially because ‘[t]he world in which we immediately live, that in which we strive, succeed, and are defeated is pre-eminently a qualitative world. What we act for, suffer, and enjoy are things in their qualitative determinations’ (Dewey, 1931: 93). These qualitative determinations are ‘felt’, hence the importance of affects. We need only think of the many value-related adjectives that are emotionally charged or denote characteristic emotions – whether they are apparently neutral adjectives (good, beautiful, pleasant, honest, charitable, cruel, etc.) or express affects more explicitly (moving, exalting, disgusting, repugnant, disturbing, horrible, frightening, etc).

It is by these emotions that we realise the things that are dear to us, or important to us, things we cherish or to which we are attached; or again the real content of our momentary orientations, our concerns, preferences, objectives and expectations. This is so because our emotions are aroused by a differential between on the one hand what is happening, and on the other our orientations, preferences, expectations, values and so on. In the words of Pierre Livet (2002), emotions are the ‘sensors of our values’ (sondes de nos valeurs). And, in most cases, it is by sharing our emotions that we are assured of the common, shared nature of our collective reference points, which include these values. However, apart from this revelation function, the emotions also accomplish real work. It is emotions that operate the selections and rejections mentioned earlier.

The idea of ‘emotional work’ is not absent from psychology and sociology. Thus, in psychology, in approaching the emotions in terms of ‘appraisal’, emotional work is defined in terms of the immediate appreciation and evaluation of situations, and the production of rapid adjustments. People assess emotionally what happens to them and what they are faced with from the viewpoint of the implications for their personal well-being and respond without thinking. In sociology, the idea of ‘emotional work’ was introduced by Arlie Hochschild (1979), in a Goffmanian perspective of managing ‘face’. But this can be taken further: we may consider that emotional work does not consist solely of appreciating and assessing situations and producing rapid responses, but also of determining the pertinent elements of a situation in order to deal with that situation, in order to unify the diverse aspects of the experience, more particularly by linking its different components (intellectual, practical, affective).

Are Values Principles?

One of the meanings Nathalie Heinich confers on the word ‘value’ is that of a ‘principle for evaluation’: ‘beauty’, ‘responsibility’, ‘morality’, ‘authenticity’, ‘fidelity’, ‘generosity’, ‘effectiveness’, ‘decency’, etc. could be deemed to be such ‘principle values’. She claims that they function as the ‘causes’ of evaluations. Characteristically, they are ‘autotelic’, that is, they ‘do not rest on anything other than themselves’ (ne reposent sur rien d’autre qu’elles-mêmes), being ‘in themselves their own goal and their own necessity’ (à elles-mêmes leur propre but et leur propre nécessité), and put an end to ‘the series of “why”s that justify the action’ (la série des ‘pourquoi’ justificateurs de l’action) (Heinich, 2017b: 299). This characterisation is very interesting, but it needs to be discussed. Let us take the case of ‘beauty’: beauty is both a way of reaching a qualification in terms of value, an evaluation criterion, and probably also a form of an ideal. Is it also a principle with causal power? I very much doubt it. To justify my scepticism, I shall consider a number of points.

It has often been noted that there is a risk of hypostasis in using an abstract term such as ‘beauty’ (although it is not because the word is abstract that the value it designates is also abstract): on the basis of emotional ecstasy (this is a particular emotion) provoked by the capacity of an object to arouse admiration, beauty acquires, by hypostatisation, the status of a pure substance, or even the status of an ‘essence of intuition’ (Dewey, 2005 [1934a]: 162). To avoid the error inherent in such a conversion of an emotional term into a substance or an essence, a tendency found in any intellectualised restitution of an immediate experience that is firstly qualitative, it would be better to start out from the adjective (‘that’s beautiful!’) that expresses an emotion. And when we use the word ‘beauty’ to designate ‘the total esthetic quality of an experience’, it is recommended that we ‘deal with the experience itself’ and understand whence and how the quality proceeds; in that case, ‘beauty is the response to that which to reflection is the consummated movement of matter integrated through its inner relations into a single qualitative whole’ (Dewey, 2005 [1934a]: 162; my italics).

A priori, an emotional term does not denote an evaluation principle. But what exactly is such a principle? We may begin to elucidate these questions by distinguishing between standards, principles and criteria. Certain evaluations are related to measurement; they therefore require standards or benchmarks. A standard may be considered independently of the function it exercises in an evaluation. This is true, for example, of the Munsell colour charts used by archaeologists to differentiate the colours of the soil they are investigating: they contain numerous colour samples. In one sense, the samples are never anything more than colours printed onto paper and a certain number of things may be said about their differences and their similarities. But when they are compared with the colours of the soil that the archaeologist wants to identify, they take on the value of a benchmark and become a tool for qualification and justification. We need standards of this type, i.e. external, public invariants, determined by convention, to be able to identify things, qualify them, differentiate them, compare them, etc. But they are only tools when they are actively used for measurement purposes. In reality, standards are not so much values as rules (although they may indeed have value), and the main operation underlying their use is comparison. A ‘principle-value’ such as beauty does not seem to be of this order, since it is not sufficiently determined, and it is not truly prescriptive.

In fact, not all standards are rules. They may also embody relatively vague reference ideals. The different practices (in trades, medicine, science and technology, the arts, for instance) include standards of excellence that are inherent to them; committing to them means deferring to the standards, which then become not only criteria for the evaluation of one’s own performances and those of others, but also markers for developing one’s competencies and guiding the exercise of one’s abilities (McIntyre, 1984 [1981]). These standards are, however, different from rules because they are not prescriptive and the ideals they embody are rather vague.

It is not the type of mediation provided by standards or benchmarks that seems to be involved in value judgements. Some evaluations are measurements, but generally speaking judging is not the same as measuring. Measuring means comparing one thing with another in order to determine a value or attribute a price (a given object has a particular length, weighs so much, is so big, is such a colour, is worth so much, etc.) or in order to carry out a specific operation, whereas judging means apprehending the qualities of an individual object by interacting with it, more or less thoroughly. The object has to penetrate the experience of the person doing the judging ‘by interaction with his own sensitivity and his knowledge and funded store from past experiences’ (Dewey, 2005 [1934a]: 357). Recourse to a predetermined external, public rule may have only a very subordinate role in this. It is true that an appreciation is made with regard to values, but the appreciated qualities are not ‘categorize[d] . . . as values’ (Dewey, 2005 [1934a]: 356).

Judgement requires criteria rather than standards and benchmarks; the criteria are not external and they are in no way prescriptive; they serve as markers for the qualitative appreciation of the various aspects of an object. To appreciate the object it is necessary to perceive it in a certain way or from a particular viewpoint – for example in terms of its composition, structure, coherence, homogeneity, harmony, conjunction of colours, incorporation of its parts in the whole, its congruence with its surroundings, etc. Anyone afflicted with ‘aspect blindness’ (inability to grasp a generally organised shape or thing, or the pertinent traits of a composition) will not be able to appreciate an object qualitatively. In general, qualitative appreciation is multi-criterial, requiring abilities and techniques – skill, in a word. Examples include knowing how to taste wine and appreciate a musical performance.

Now, are the criteria for judgement principles? It all depends on what is meant by ‘principle’. Principles are evoked in logic, in mathematics, in enquiry theory, in moral philosophy, and in ordinary discussions on the soundness of a course of conduct. In the philosophical tradition, there has always been a tendency to envisage principles as rigid things, determined a priori and exterior to the actual evaluation processes. The pragmatists have criticised this propensity and envisaged principles rather like habits, methodical rules, or hypotheses. In logic, principles formulate habits of inference that produce stable conclusions in their reasonings (as Peirce argues). In enquiry theory, they formulate operations to be performed, as well as markers for evaluating the conduct of the enquiry in the context of its accomplishment (as Dewey argues). In ethics, principles are revisable probabilistic generalisations with no normative element that we use in our moral judgements. They formulate ways of responding to a certain type of situation, but they derive their authority from what the situation imposes and not from some intrinsic nature. In fact it most often involves generalisations based on earlier empirical enquiries carried out in problematic situations. These generalisations provide points of view and hypotheses for deliberating on the spot. As hypotheses, principles are tested, confirmed, and altered according to the results and consequences (‘valued’ and evaluated) that occur when they are used as a basis for action.

Do values comply with this definition? Partly, yes, and that is how we are able to assure ourselves of their objectivity, i.e. by taking them as foundations for action and by seeing whether we ‘value’ the results: ‘Try, and you’ll see if you really appreciate it. If you do, you will be able to rely on it in the future.’ By putting value judgements into practice, we obtain data enabling us to test their validity.

Values can therefore be treated like principles. I feel, however, that they correspond more to ‘conceptions of the desirable’, to use Talcott Parsons’ definition (1968: 136). ² As such, however, they do not designate things that would be so in general or a priori. ‘The desirable’ is elaborated on the one hand in what Charles Taylor (1998 [1989]) calls ‘strong evaluations’ and on the other in the projections of the idealising imagination. A ‘strong evaluation’ takes account of real conditions, as apprehended in the processing of situations. But determination of what is ‘desirable’ also requires being able to draw distinctions between what deserves to be maintained or discarded in spontaneous ‘valuations’, between what deserves to be pursued or abandoned, and hence distinctions between good and bad, better and worse, higher and lower, etc. Such distinctions suppose being able to refer to independent criteria of desires, inclinations and choices of the moment.

Ludwig Wittgenstein noted that there is a tendency to sublimate standards, to desire them in their pure form, freed of their materiality. But this is an illusion, since it is because of their materiality that benchmarks – at least when they are physical things – are means of measurement and representation: ‘Length cannot exist without a body, and even if I understand that, in a specific sense, the only length is that of the ruler doing the measuring, what I put in my pocket is no less the ruler, the body, and not the length’ (Wittgenstein, quoted [in French] in Bouveresse, 2006: 63). To a certain extent, the same is true of ‘principle-values’. In his Remarks on Frazer’s Golden Bough, Wittgenstein notes on the subject of death and beauty: ‘Nothing is as dead as death; nothing is as beautiful as beauty itself’. The image we use to represent reality here consists of thinking of beauty, death, etc. as pure (concentrated) substances, whereas they are present as ingredients (in Bouveresse, 2006: 64). This propensity to hypostasise adjectival terms with a high emotional content is obviously very present in the conception we spontaneously have of values.

As we have said, values are most frequently designated by abstract nouns: beauty, truth, justice, democracy, equality, liberty, solidarity, human dignity, etc. It would be fair to say that these concepts have added new values, new interests and new incentives to life, as well as new feelings of admiration. Thus William James (2007 [1909]: 224) asked that we do not forget that these concepts are always to be connected to active life: ‘Persons who have certain concepts are animated otherwise, pursue their own vital careers differently. It doesn’t necessarily follow that they understand other vital careers more intimately’. William James also noted that ‘abstractness per se seems to have a touch of ideality’ and arouses a veneration that is not produced by the concrete realisations of abstract ideas, which look like very poor relations in comparison:

The veritable ‘cash-value’ of the idea seems to cleave to it only in the abstract status. Truth at large, as Royce contends . . . appears another thing altogether from the true particulars in which it is best to believe. It transcends in value all those ‘expediencies’, and is something to live for, whether expedient or inexpedient. Truth with a big T is a ‘momentous issue’; truths in detail are ‘poor scraps’, mere ‘crumbling successes’. (James 2007 [1909]: 224)

In a word, abstract ideas and abstract nouns arouse real veneration because of the ‘touch of ideality’ they transmit, i.e. because they designate ideals produced by the idealising imagination. But as we shall see, veneration is only one of the attitudes that an ideal may arouse.

Does a value ‘not rest on anything other than itself’ (ne repose-t-elle sur rien d’autre qu’elle-même) (Heinich, 2017b: 304)? Does it represent ‘a stop on argumentation’ (une butée de l’argumentation) (Heinich, 2017b: 304), in the sense that it may be sufficient to itself? The risk with such formulations is that they give credit to those who make values something ultimate, supreme, absolute. To avoid this risk, it is firstly recommended that terms such as ‘ultimate’, ‘last’, etc. should only be used in the temporal sense, as Peirce and Dewey suggested:

A thing may be ultimate in the sense of coming last in a given temporal series; so that it is ultimate for that series . . . For me the method of intelligent action is precisely such an ultimate value. It is the last, the final or closing, thing we come upon in inquiry into inquiry. But the place it occupies in the temporal manifestation of inquiry is what makes it such a value, not some property it possesses in and of itself, in the isolation of non-relatedness. It is ultimate in use and function; it does not claim to be ultimate because of an absolute ‘inherent nature’ making it sacrosanct, a transcendent object of worship. (Dewey, 1939: 77)

A value may, therefore, ultimately be less a ‘stop on argumentation’ (we shall see that values and argumentation do not mix) than a final point of the enquiry.

Another strategy for avoiding making values ultimate things is to relate them to situations as well as to contexts. We may apply to values what John Dewey said about goods always being the values of a situation: there is no good in itself or absolute good, but only ‘the goods of a situation’. Values are discovered when there is conflict, when it is necessary to choose between directions or choices, by thinking about the conditions and the consequences: ‘No one lives in a world in which he has found everything at all times perfect. If he understands the meaning of this fact he has learned to be alive to possibilities. The potential better will then be regarded as the good – and the only good – of any situation’ (Dewey, 1939: 72). These goods/values discovered in situ by reflection often in fact proceed from a re-evaluation of goods/values arising out of the immediate experience through affective-motor valuations. The same applies to criteria: rather than being a priori and determined independently from practices, they are drawn up and configured as means by their use in situ.

Once it has been ensured that the previous risk no longer obtains, we may indeed say that a value ‘does not rest on anything other than itself’. But that supposes considering values more as rather vague reference ideals than as principles. Ideals do not in any way ‘cause’ conduct; they arouse convictions and aspirations, allegiances and commitments; they indicate directions for conduct, or for improving the quality of the experience, and they make certain experiences possible, on condition, however, that they are something other than images of ‘pure substances’.

Values as Ideals

Considering values as ideals is something Émile Durkheim did. He believed that values are essentially ideal purposes, serving as forces for liaison and motivation in social life. According to Durkheim, it is through the emotional collective experiences of creative effervescence that collective life ‘reach[es] a certain degree of intensity’, that ideals are formed. He writes that idealisation is an ‘essential characteristic of religions’ (Durkheim, 2008 [1912]: 602). It takes place within rituals that are extraordinary moments of exaltation of a society’s moral life, during which the society becomes aware of itself and nurtures the feeling it has of itself. Idealisation is thus ‘a natural product of social life’ (Durkheim, 2008 [1912]: 603). Nathalie Heinich (2017b: 294) does not agree with this comparison of values as ideals with religion: ‘The religious paradigm is of no relevance to the issue of values’. ³ In fact, it all depends on the conception one has of the religious. I feel there is a different conception from that of Durkheim that may shed light on the nature of values.

For Durkheim (2008 [1912]: 602), ideals are something added to and above the real, as is the sacred: they ‘substitute for the real world another different one, to which [men] transport themselves by thought’. Added onto the real world, the ideal makes it possible to transcend ordinary experience and conceive of a different world. That is why idealisation is vital for a society: ‘A society can neither create itself nor recreate itself without . . . creating an ideal. This creation is not a sort of work of supererogation for it, by which it would complete itself, being already formed; it is the act by which it is periodically made and remade’ (Durkheim, 2008 [1912]: 602).

For Durkheim (1970: 265), values have an authority, which comes largely from their sacralisation and places them ‘outside and above all temporal interests’, arousing fear of and respect for them as well as love and attraction (because they idealise and depict a new life).

The individual or the human person represents such a value in modern societies: the dignity of the individual has become the ‘religion of humanity’ and the individual has become the object of a cult of which he/she is ‘both the object and the agent’:

The human person, whose definition is like the touchstone according to which Good must be distinguished from Evil, is considered sacred, in the ritual sense of the word, so to speak. It has something of the transcendent majesty that the Churches of all times attribute to their Gods, it is conceived as invested with this mysterious property that empties around holy things, that abstracts them from vulgar contacts and detracts them from the common flow. And that is precisely where the respect for it comes from. Whoever makes an attempt on a man’s life, a man’s freedom, a man’s honour, inspires a sense of horror in us, in every way similar to that experienced by the believer who sees his idol profaned. Such a morality is therefore not simply a hygienic discipline or a wise strategy of existence; it is a religion in which Man is both the faithful and the God. (Durkheim, 1970: 264–265)

Durkheim goes on to explain how the individual has become a god for humans and how ‘the religion of humanity’ has become today’s religion. This new religion has all the classic attributes of a religion: the human person is the only thing which ‘collective sensitivity [could] now be taken from’ (Durkheim, 1970: 271) and ensures ‘the communion of spirits’:

As soon as an end is pursued by an entire people, it acquires, as a result of this unanimous accession, a kind of moral supremacy that elevates it far above private ends and thus gives it a religious character. On the other hand, it is obvious that a society cannot be coherent if there is no intellectual and moral community among its members. (Durkheim, 1970: 270–271)

Hans Joas has devoted an entire book to this sacralisation of the person and to the genealogy of human rights. He writes that the genesis of values is a contingent process and that ‘our values are historical individualities’ (Joas, 2013: 3). They can only guide behaviour if they are upheld by institutions and incorporated into practices. What must be explained above all is how our commitments to values (value-commitments) are formed: ‘Our commitment to values and our notion of what is valuable emerge from experiences and our processing of them; this shows them to be contingent rather than necessary’ (Joas, 2013: 3).

Joas emphasises that one of the characteristics of value-commitments is that they have an affective component. This is why we treat having values differently from having opinions or beliefs; it is not because of a recognition of their cognitive validity that we adhere to them (which would strongly imply the criterion of the ‘stop on argumentation’ that Nathalie Heinich adopts). Their power of conviction is not that of discursive persuasion achieved by argument, but the specific power of an ideal. Consequently, communication on the subject of values cannot be reduced to an exchange of rational arguments. Hans Joas recalls more particularly that it is not possible to discuss values in an ‘atomistic’ fashion, as if they were discrete, autonomous opinions; they form groups or clusters, in which they are entwined.

Unlike Durkheim, Joas does not consider the sacralisation of the human person to be a phenomenon of the same type as religion. Nevertheless, in as much as it is in the order of an ideal, it does have something to do with the religious sphere. This is a point of view that was to be found earlier in the work of Dewey, more particularly in his essay on the religious, entitled A Common Faith (2011 [1934b]). For Dewey, what the religious connotes is not acceptance of the beliefs and practices of an established religion, but a practical attitude, which is embodied by the choice of a way of life based on an allegiance to ideals: ‘I should describe [religious] faith as the unification of the self through allegiance to inclusive ideal ends, which imagination presents to us and to which the human will responds as worthy of controlling our desires and choices’ (Dewey, 2011 [1934b]: 24). This allegiance does indeed have an affective element, but it is essentially a practical orientation, imposing demands and involving an effort to change oneself and to change the world by one’s practices. It stems from moral conviction, which is not the same thing as intellectual assent to beliefs and arguments:

Conviction in the moral sense signifies being conquered, vanquished, in our active nature by an ideal end; it signifies acknowledgment of its rightful claim over our desires and purposes. Such acknowledgment is practical, not primarily intellectual . . . The authority of an ideal over choice and conduct is the authority of an ideal, not of a fact, of a truth guaranteed to intellect, not of the status of the one who propounds the truth (Dewey, 2011 [1934b]: 106).

The attitudes that constitute the religious are therefore nothing extraordinary or irrational; they belong to the dynamic of the natural experience of humans, i.e. the formation and unification, via the creative imagination, of ideal purposes or ideals with authority – in a word, values. These values are extracted from human associations by grasping the real possibilities in experience: this involves projections on the basis of goods that are concretely experienced in family relationships, in neighbourhood and workplace relationships, in the exercise of citizenship, in the practice of arts, or in science. But people have always tended to substantiate them and have always tried to back their authority by that of a supernatural being or invisible forces. We have hypostatised the ideal purposes projected and unified by the imagination into an external, antecedent existence. But ideals do not need the support of such forces, since they are not divorced from ordinary experience; they need any other support than the habits, practices and significations that animate them:

[W]hat I have tried to show is that the ideal itself has its roots in natural conditions; it emerges when the imagination idealizes existence by laying hold of the possibilities offered to thought and action. There are values, goods, actually realized upon a natural basis – the goods of human association, of art and knowledge. The idealizing imagination seizes upon the most precious things found in the climacteric moments of experience and projects them. We need no external criterion and guarantee for their goodness. They are had, they exist as good, and out of them we frame our ideal ends. (Dewey, 2011 [1934b]: 137–138)

But it is not because they are produced by the imagination that these ideals are made out of ‘imaginary stuff’: ‘They are made out of the hard stuff of the world of physical and social experience’ (Dewey, 2011 [1934b]: 139).

These ideals are used as evaluation tools, means of guiding conduct, and foundations for criticism: ‘They supply the meanings in terms of which life is judged, esteemed, and criticized’ (Dewey, 2012 [1925]: 193). They make it possible to evaluate and criticise that which exists, and create an awareness of what is missing, because they generate ‘a sense of possibilities’ by which some of the negative aspects of reality are discovered:

A sense of possibilities that are unrealized and that might be realized are when they are put in contrast with actual conditions, the most penetrating ‘criticism’ of the latter that can be made. It is by a sense of possibilities opening before us that we become aware of constrictions that hem us in and of burdens that oppress. (Dewey, 2005 [1934a]: 396)

We are not in the register of causality here.

Dewey states, however, that the fact of a moral faith having ideal purposes does not necessarily make it religious. Via emotion, faith must produce a unification of the self and its harmonisation with the universe:

The religious is ‘morality touched by emotion’ only when the ends of moral conviction arouse emotions that are not only intense but are actuated and supported by ends so inclusive that they unify the self. The inclusiveness of the end in relation to both self and the ‘universe’ to which an inclusive self is related is indispensable. (Dewey, 2005 [1934a]: 108)

In a sense, Dewey agrees with Durkheim’s definition of values in terms of ideals with a specific form of authority of their own that is of a moral nature. But since, with George H. Mead, he rejects the idea of society being a particular psychic being ‘of a new kind’ (Durkheim) – society is not mental; the spirit, the imagination, the conscience are not where society takes place, George H. Mead repeats over and over again – in the same way as he refuses an immediate link between the religious and the sacred or the supernatural, and rejects the explanation of the organisation of social life by sui generis collective forces (and more particularly collective representations), he can be more economical and more empirical than Durkheim in his analysis of the social genesis of values. They correspond to the projections of the idealising imagination, taking up the possibilities discovered in real life. Thus it is the experience itself that reveals the values that in turn make new experiences possible: ‘The value of experience is not only in the ideals it reveals, but in its power to disclose many ideals, a power more germinal and more significant than any revealed ideal, since it includes them in its stride, shatters and remakes them. One may even reverse the statement and say the value of ideals lies in the experiences to which they lead’ (Dewey, 2005 [1934a]: 370).

At the origin of the allegiance that values arouse there is therefore not the sacralisation of ideals, as with Durkheim, nor the capture of the common affect generated by affective imitation (as with Spinoza, taken up by André Orléan [2011]), but on the one hand the attachment implied by moral conviction for ideal purposes discovered in experience and on the other the shared emotions produced by exploration, in art, in communication and in social enquiry, of real possibilities not yet realised. Emotions are important, but it is not emotions that are the source of values and their authority: it is the imagination and its anchorage in real life of the possibilities that the imagination discovers and projects. It is the ideal purposes of moral conviction that produce emotions, rather than the other way around.

The Authority of Values

It has frequently been argued that values cannot be universalised like norms. This is because they are only subjective preferences, or possibly collective cultural preferences (freedom, equality and fraternity in a republic, for example), or goods that are desired by all. If values are no more than ‘intersubjectively shared preferences’ (Habermas, 1997 [1992]: 278), and if value judgements do no more than establish such preferences, by weighting various criteria, then the question of the legitimacy and authority of values does not arise; authority is replaced by the mere attraction of goods. Goods are commonly thought to deserve the joint effort of achieving them. It may be conceded that values cannot be considered as being universal, but if they do impose themselves as the final stage in enquiry (as ‘the method of intelligent action’ in the foregoing quotation from Dewey), they are in a sense universalisable.

We have seen that there is something like an allegiance to values that is not the same as an attraction to preferences: allegiance is in the order of a submission to an authority, which supposes that values may impose themselves as legitimate things that are valid for all. In L’Empire de la Valeur (English title: The Empire of Value), André Orléan (2011) explains this authority as a capture of the ‘power of the multitude’, a power that is produced by the interactions of individuals imitating and affecting each other in the formation of their desires. This type of explanation is based on a fiction that is constantly being reproduced by conventionalists such as Orléan, who claim to operate a methodological individualism. But this is not acceptable: individuals are always already plunged into a form of life and a social life where meaning is instituted, which is not entirely the same thing as being ‘plunged into interactions’ (plongés dans les interactions), as Orléan puts it. They are sensitive to certain values and acknowledge their authority because, through the customs and institutions of their social milieu, they have developed certain practices that they ‘value’. The dependence of values on ‘valued’ instituted practices forces us to relativise Nathalie Heinich’s argument that values ‘do not rest on anything other than themselves’. This primacy of practices, if it is proven, also goes against the importance afforded to ‘axiological representations’. However, it also forces us to invoke a different mechanism than the one adopted by Dewey (a specific type of conviction connected with ideals).

What is this mechanism? Orléan’s explanation has the merit of making the authority of values dependent on a collective dynamic. It is precisely this that it would be better to identify. We have seen that Durkheim believed in the ability of rituals to create idea-values by producing a collective effervescence in which moral life is exalted. I shall propose a different type of explanation starting out from noting that, in interactions and in practical actions, individuals suppose that the markers and criteria they use are shared – they are therefore merely putative or ‘virtual’ (virtuels), to use the language of Pierre Livet (1994). Their effective sharing and recognition are never assured immediately, and only real interactions can lift the uncertainty, at least in part. Such uncertainty is particularly true for values, at least for those that involve personal commitment: I am never immediately assured that other people care about the same things as I do, have the same hierarchy of importance that I have, or venerate what I venerate. This uncertainty can be a source of emotion, as can its reduction. Indeed, sharing emotions is often a means of making sure that collective markers – and particularly values – are shared and recognised by the majority (rather than by all) as being authoritative for them. Discovering that others do not respect them may conversely give rise to negative and often unpleasant affects, and cause uncertainty to well up regarding the extent of their sharing, since the uncertainty engenders additional emotions (Livet, 2002).

Social interactions nevertheless have an effective mechanism for reducing uncertainty. As Niklas Luhmann explains (1968: 24), trust ‘serves to overcome an element of uncertainty in the behaviour of other people which is experienced as the unpredictability of change in an object’. But there are different types of trust (Quéré, 2017). One type plays a fundamental role in social life: we may call it Ur-trust. It is a type of primary trust, consisting of relying on others with no hesitation or thought, and without first carrying out checks and evaluations. This primary trust is a practical attitude, i.e. a course of action. It forms the basis for the coordination of most of our everyday activities, in which commitments are made without hesitation or thought. Everyone spontaneously relies on everyone else when taking as the basis for inference and action the instituted frameworks of social life, i.e. established usages, habits and social significations, or again what everyone is supposed to know, or consider obvious with regard to beliefs and values about social life, which are deemed legitimate. But nothing guarantees they will coordinate effectively, and it is interactions that serve as tests.

This is what Harold Garfinkel (2007 [1967]) showed in his breaching experiments, by adopting the vocabulary of Alfred Schütz’s social phenomenology: when they act and interact, social agents rely on each other to adopt a commitment in favour of ‘a legitimate order of beliefs about life in society’, i.e. allegiance to custom, to the institutions of social life, and to instituted significations and values. In Garfinkel’s phenomenological vocabulary, they rely on each other to subscribe to the ‘natural facts of life in society’ as morality, to refer to the ‘institutionalised characteristics’ of the group, and to do so in a truly specific posture, namely a morally motivated attachment, which as Durkheim explained combines submission to what is compulsory and attraction to what is considered good or legitimate – in short, a form of moral conviction. In other words, ‘bona fide’ society’s members normatively expect each other to refer to what they suppose is common to them all:

To social structures which [they] treated as actually or potentially known in common . . . And then, not to any social structures known in common, but to normatively valued social structures which the subject[s] accepted as conditions that [their] decisions, with respect to [their] own sensible and realistic grasp of [their] circumstances and the “good” character of the . . . advice [of others], had to satisfy. These social structures consisted of normative features of the social system seen from within, which, for the subject[s], were definitive of [their] memberships in the various collectivities that were referred to. (Garfinkel (2007 [1967]: 172)

Garfinkel implicitly takes up here the Roman conception of mores, as customs that are ‘valued’ and are authoritative because they are considered as serving the common good.

Social agents, Garfinkel also explains, apprehend the axiological and moral order of collective life as an objective, independent of them – an objectivity and an independence they produce without actually realising they do it. And it is by honouring, in a justified fashion, i.e. as a matter of thoughtless moral conviction, the ‘background expectancies of everyday life’ that they produce the characteristics of real society, particularly the transcendence of norms and values. When an individual discovers, in the course of interaction, that his/her partner(s) do not meet these expectations, he/she may be seriously disoriented, and feel negative affects, possibly even including towards him-/herself; in most cases, he/she makes every effort to ‘normalise’ the situation (Garfinkel (2007 [1967], chap. 2).

From this point of view, we are therefore able to say that the authority that values have is supported by both the attitudes that people adopt in carrying out their interactions, and the immediate operations they carry out in respect of each other: hypothetical attribution of knowledge and beliefs, supposition of shared collective markers, conjecture of a common commitment to the ‘natural facts of life in society’, or in favour of beliefs and values that are considered to be legitimate, implicit verification of these commitments in the interactions, etc. These attitudes include the ‘valuation’ not only of ‘social structures’ (in Garfinkel’s language), but also of the customs or instituted practices, in as much as they are driven by certain values and significations.

I am not sure that this explanation entirely avoids the ‘sociologist’s fallacy’ mentioned at the beginning. It is nevertheless enlightening. What Garfinkel calls a ‘legitimate order of beliefs about life in society’ pretty well matches what Taylor (2004) calls a ‘social imaginary’. This is a socio-historical creation; there are specifically ‘modern’ social imaginaries (for example, the idea of public space, or the sovereignty of a people). It involves the way in which people create an image of their social existence, of the way they ensure their coexistence, the way things happen or should happen between them and their fellow-citizens, etc. It is not merely a matter of apprehending the norms and values that underpin social interactions, but also a broader conception of what is and what is not ‘normal’ (social normativity includes ‘normality’), of what is and what is not valuable, of what is and what is not fair, as well as ideas of what is desirable, what should be, the way we should live together in society, and why. Thus, these moral images of living in society make way for aspirations and ideals. A social imaginary supplies paradigms for hierarchising in terms of their value or whether they deserve to be pursued, modes of behaviour, modes of living, ways of interacting with others, and modes of government or management of public affairs.

Does that mean these are mental representations? Nathalie Heinich (2017b: 311) thinks there are ‘axiological representations’ of a mental nature, i.e. representations ‘which humans make for themselves about what is worthy of being appreciated or praised, in what way they cannot be reduced to facts of nature’ (que se font les humains de ce qui est digne d’être apprécié ou loué, ce en quoi elles sont irréductibles à des faits de nature). These representations, she says, are cultural and not natural; they form a system and they are shared. This characterisation ties in with our earlier definition of the social imaginary, but it is open to doubt whether it is as a set of representations or beliefs that this imaginary guides behaviours and practices, and inspires commitment and aspirations. It only remains for me to take one final step into the ‘pragmatisation’ of the field of value that Nathalie Heinich proposes.

Conclusion

It is impossible to dissociate from the idea of representation the idea of a subject of representations, with a mind to house its representations and carry out operations on them. We do indeed form representations, but the larger part of our conduct and our actions is based less on representations than on an unformulated, and partially unformulatable, apprehension of things, or on a tacit apprehension of our overall situation. This apprehension is firstly unformulatable in that it can never be expressed adequately in statements because of its unlimited and indefinite nature (Taylor, 1995).

It is also unformulatable because it is the practices and institutions of social life, and not what is inside people’s heads, that are the prime locations for significations, norms and values. Usage, social habits and instituted practices incorporate – as implicit constitutive elements – ideas, norms, standards and ideals that order them, make them practicable and give them a meaning that give rise to commitments (as for example voting in a democratic regime). It is in this incorporated form that values first operate. They may indeed be formulated or represented, as are rules and norms. But it is not primarily as such that they guide behaviours; they are immanent to practices. That is why, if we are sensitive to certain values, it is because, via usages and institutions that are contingent, we have developed certain practices, which are socially ‘valued’. Thus in the modern era, polite conversation and persuasive discussion, on condition that the participants are on a quasi-equal footing, have replaced not only ritualised combat and medieval jousting but also the exchanges between individuals belonging to a hierarchical order.

‘In the beginning was the deed’, wrote Wittgenstein, echoing Goethe. In doing so he noted that our ethical concepts, in particular our sense of human dignity, are anchored in what we do, particularly in the way in which we relate specifically to our bodies, and in the habitual way in which we treat each other in our interactions. One of his notes in the 1930s demonstrates this anchoring:

Mutilate completely a man, cut off his arms and legs, nose and ears, and then see what remains of his self-respect and his dignity, and to what point his concepts of these things are still the same. We don’t suspect at all, how these concepts depend on the habitual, normal state of our bodies. What would happen to them if we were led by a leash attached to a ring through our tongues? How much then still remains of a man in him? Into what state does such a man sink? We don’t know that we are standing on a high narrow rock and surrounded by precipices, in which everything looks different. (quoted in Shusterman, 2002: 103–104)

If we transpose this note to the matter of values, we may say that they are not separate from a socially ‘valued’ ‘habitual, normal state’ of interactions and social relations, and more broadly of modes of coexistence. This state may also possibly be criticised for its failings and shortcomings, in the light of the real possibilities revealed by experience – that which is ‘normal’ is indeed conjugated not only in the indicative but also in the conditional (‘what would be normal’ would be to . . . or . . .). But it is above all contingent. . . and ‘surrounded by precipices’!