Narratives of trust: sharing knowledge as a second-order emotion

Narratives of our emotions

Emotions tend to resemble appear as complex structures which include more than one emotion at the same time or in succession in very close temporal proximity (Barbalet, 1998; Benski, 2005; Collins, 1990; Flam, 2005; Scheff, 2009; Yang, 2000). Benski (2011)suggests adopting the concept of ‘constellations of emotions’ to deal with the complex nature of emotional dimensions, where situations give rise to emotional experiences which often cannot be adequately accounted for by focusing on a single emotion. These constellations are composed of similar emotions, which work in a similar direction.

In studying this complex panorama, language can help us to discover these emotions. They are located in texts and have narrative structures, as we will show in this section. In these narratives, emotions can take any number of different turns (Stets, 2010), moving from positive to negative, and back again from negative to positive. We will now explain and predict the particular sequence and type of emotions that can ensue in encounters and narratives. Williamson (2011) speaks of ‘chains of emotion’ because they are sequenced in our discourses. Narratives are constructs that are woven in societal history and highlight certain aspects of society in our lives. Narratives are instruments for understanding emotions. They have two different roles: one is to describe a situation; the other is to justify what we learn and what happens in our lives. Hutto (2008) claims that narratives make sense of intentional actions in our lives. We learn about our emotions thanks to our narratives from the time we are children to the present. De Sousa’s concept (1987) of paradigm scenarios as experiences that shape one’s learning of different emotions might be useful to help us understand the previous statement. Paradigm scenarios involve the characteristic objects of a specific emotion-type and the response to a specific situation, determined by a mix of biological and cultural factors.

Following Harré (2009), the grammatical subject of most words used to describe emotion is the person, because these are used to convey the idea that people are angry, sad, proud, chagrined, joyful, and so on. So people’s narratives are composed of an emotion-grammar to describe emotions. In emotion-grammar, people make use of concepts such as ‘love’, ‘jealousy’, ‘trust’, and so on. This type of grammar changes in every context and includes semantic and syntactic rules and storylines. Thanks to the work of philosophers and psychologists, we now have concepts to explain these emotion processes or constructions—narratives. These narratives allow to us to understand how phrases such as “being angry,” “being jealous”, and “being ashamed” are compared. Language is a tool of emotional life, which helps to modify and solidify emotional activity (Mendonça, 2008). It is also a creative tool for handling emotional difficulties in our lives.

According to Goldie (2012), people do not feel the same emotion over time, because time changes, the context changes, and we observe emotions through an external perspective of what happened. Goldie (2012) introduces the internal/external standpoint, which we can observe in this narrative: Last week that person said important and friendly things to me I trust him, and he trusts me to confide in me. Today, when I remember that event, I discover that I love him, and not just trust him.

In this person’s external standpoint narrative (today about an event which happened the last week), s/he discovers an internal emotion, love. A sort of emotional gap, or a retrospective emotion that s/he did not have at the time (last week), or which was probably caused by this situation of trust. This internal/external standpoint is an action that involves rethinking the past to change the present. In emotional narratives we often go through this action: something happens in the past, and it takes on a different meaning when we re-elaborate on it in the present. Trust appears like a second-order emotion and love like a first-order emotion in the previous narrative. But what is the difference between a first and a second-order emotion?

To understand this difference, we need to introduce the notion of forms of sociation, elaborated by Georg Simmel in 1908 (1992). Sociologists consider special forms of sociation (observable object of analysis in time and in space) to exist that persons interiorize during socialization. These forms allow us to communicate, understand and be understood in society, as Natàlia Cantó-Milà (2012) clarifies. But many of these do not appear as autonomous entities in our daily lives and relationships because they are analytical abstractions used to achieve the strictly social within social relationship (Cantó-Milà, 2012, p.12). For this reason, Simmel (1992) introduces the concept of forms of second order. For him, gratitude and faithfulness are second-order forms because they are forms of forms: “instruments of relations which already exist and endure”, related to the first-order forms to the “material contents and motives of social life” (Simmel, 1964, p. 379). Cantó-Milà underlines that “the special nature of these second-order forms is expressed by the way in which they help to link first-order forms of sociation to the duration/durability of society” (2012, p.12).

This condition of durability can also be observed in Goldie’s work (2012): “When you grieve, you often look back on the past, on your time with the person you loved, knowing now what you did not know then: that the person you loved is now dead” (2012, p. 65). In this example, grief is a first-order emotion at the moment that describes the person’s emotions in a narrative, and love becomes a second-order emotion. Love is the basis of this new emotion of grief, while in previous examples trust is the basis of the new emotion of love. It is always an emotion about an emotional process. In the first example love is the first-order emotion, but in Goldie’s example, love is a second-order emotion. Second-order emotions, as we will see later, assume different roles depending on the pattern and situation.

Cantó-Milà (2012) also argues that second-order forms shape memories, because they mark and reinterpret what is ‘memorable’ in past experiences, and connect us to our actions and the emotions that we will have in the future. Thanks to these second-order forms we can analyze the complexity of emotions, fitting them into the “facts” (McIsaac & Eich, 2004). These facts are “structured episodes” and are composed of elements of emotional experience: thought, feeling, change, expression, and so forth (Goldie, 2012). Thanks to these facts, or episodes, we have experienced emotions in a particular circumstance and confer an extension in time to momentary social bonds.

Second-order emotion and trust

For Simmel (1964, in Cantó-Milà, 2012), in most cases faithfulness and gratitude are second-order emotions because they represent a powerful cement for social relations, becoming the memory of other emotions like surprise, pleasure, joy, astonishment, etc., and linking people strongly to their society:

[I]f I am grateful and faithful to a person, to an imagined nation, to an institution, to a memory, I will normally not feel, at the same time, as if I have no place in that relationship, and no role to play in its future, either (Cantó-Milà, 2012, p.14).

Second-order emotions such as gratitude and faithfulness feature among the basic processes of sociation that have to be learned and incorporated during human experiences. Durability is a necessary aspect of our narratives, and it is important to consider the nature of this. According to Roberts (2013), “a personal relationship is a disposition of both parties to think, act, and feel in ways characteristic of the (good or bad) relationship” (p.140). One can gain or lose trust in another person, which modifies the whole emotional process and distorts it.

Second-order emotional practices have an important role in emotion management, considering that emotion management is not the conclusion of a process, but is a phase of a continuing cycle of activities (Barbalet, 2011). Emotion management is an internal process employed in the emotion regulation process, which helps modulate emotions (Sheffield Morris, Silk, Steinberg, Myers, & Robinson, 2007). Second-order emotional practices are a tool (Jakupcak, 2003), “an instrument of freedom rather than a tool of self-oppression” (De Sousa, 1990, p. 446), that enables us to cease thinking of we cease to think of our emotions as inevitable and to view them as open to modification. A person may ‘regulate’ jealousy or anger at her partner, build trust, promote rewarding actions, share knowledge and information, etc. Trust can generate a constellation of emotions, from anger to love, from jealousy to resentment in a personal relationship.

For Barbalet (2011), emotions can be regulated in an implicitly social way and explicitly through processes of self-monitoring. Such emotions can only be regulated interactionally, and so require cooperation among individuals in trust relations. In managing jealousy, a person may experience a feeling of trust in her partner or society. The regulation of a person’s jealousy draws on other emotions, such as anger or love, and is combined by other second-order emotions such as sincerity, sympathy or blame.

Of course, we need to differentiate between different situations where loss of trust arises from a rupture in a trust situation. In the first case, we have a low level of trust in someone, which can be recovered, probably through a long and hard process, but it is recoverable: My friend violated the norms of our friendship by doing something wrong to me. I will try to explain to her which she did to our relationship, and in the future she needs to avoid a repeat of this action to maintain our friendship. In the second case it is impossible to recover this trust: The same friend has repeated a negative behavior towards me several times, and we have broken off our friendship. In these two examples, we can understand how trust has a central role in our relationships.

In domestic violence, the abuser expresses anger against his partner, but this anger is (often) the result of jealousy where there is no trust in the partner. So trust in this episode is the second-order emotion and jealousy and anger are the first-order emotions. The abuser cannot manage his violent behavior, and he cannot control his jealousy. Anger management therapy could help the abuser to control his jealousy by working on trust processes with his partner. The narratives generated by the abuser have two functions: the first is that they describe how his anger works in his life, and the other is that they justify this emotion. In therapy he is asked to explain why it is important to him to express anger in a given situation and what the causes of it are. The narrative has a normative function that justifies his practice.

As researchers, we apply linguistic labels in order to recognize emotions (Jasper, 2011). When people label their own emotions, these labels begin to give their emotions shape and direction (Barrett, 2006). Mendonça (2013) states that when we have an experience we do not label the emotions according to level; when we look back at what happened, we are capable of placing the second-order emotional level of a particular emotional occurrence, but when we experience a certain emotional situation, we do not mention that second-order emotion. We do not use this type of description in narratives: This person feels love for him, because in this ‘action’ of feeling love he feels different second-order emotions, and trust is one of them. To love someone is just one first-order emotion, the emotion of love, but more actions are probably necessary to perform just one action. We need to trust that person, we need to put our life in her hands, we need to accept this loss of individuality or autonomy, etc. (and vice versa).

The complexity of studying emotions can also be seen in the following example, where the action of loving someone is composed of different second-order emotions. In Harré’s (2009) example of jealousy, A and B have a pre-emptive right to a particular good, say X (trust). In this episode, A and B probably experienced anger, or a fear of loss, or failure to get X between them. B feels an unpleasant combination of bodily feelings, including tearfulness, anxiety and so on. B does not trust A because she does not recognize the legitimacy of A to have X. And jealousy appears in the relationship between A and B out of this lack of trust. Trust works as a second-order emotion in this example. People rarely express fear, anger, jealousy, chagrin, joy, and so on, by using the corresponding words in self-description (Harré, 2009). An angry person might verbally show anger by shouting “F*** you!!!” but not “I am angry with you” without turning red in the face. First- and second-order emotions in our narratives emerge in multiples ways, rarely using the corresponding words.

Spiraling out of first-order emotion, there are all sorts of second-order emotions which depend on tacit knowledge of the first-order emotions. Harré (2009) suggests that we analyze what words are commonly used to express emotions. What do the uses of the words “jealousy” and “anger” have in common? What about “jealousy” and “love”? Second-order emotions help us answer these questions, analyzing what words and expressions have in common in the context of these emotions. Anger and jealousy are expressed when the bodily component of the emotional occasion is unpleasant. But jealousy and love are different experiences. Jealousy stands as the polar opposite of trust. The latter is a prime ingredient for describing one’s emotional experience as love for someone. The first-order emotion of love is always a kind of process where this emotional process is composed by second-order emotions. The only way to observe these differences is to analyze narratives. In empirical research, narratives represent the only way to access these emotions.

Epistemic trust as a form of sharing knowledge

We start by assuming that thanks to communication, we obtain and offer information. Speech is the action of sharing information, which we count as knowledge (Faulkner, 2000; McDowell, 1994). However, to trust a person’s words, we must believe that she offers true information. We regard “testimony” as “presentation-as-true”. When a speaker presents her testimony as true, the listener is willing to accept it in a trust relationship.

To understand the process of trust as second-order emotion better, we will use the role of testimony (Broncano, 2008). Testimony acts as a kind of social bond, created by the mutual intention of sharing knowledge. For Craig (1990), trusting others means sharing knowledge, and Origgi (2004) further adds that every testimonial situation involves strong bonds of trust between the speaker and the listener. In legal courts, testimonies are accepted as information that may or may not be trustworthy, but it is for the jury and judge to make that decision. Trust is the basis of this relationship; speaker and listener share the same second-order emotion. We agree with Origgi that trust goes deeper than is supposed by mere inference from the reliability of others’ beliefs and the probability of them being true given the history of the speaker. Origgi (2004) postulates a sort of epistemic trust, according to which the speaker’s commitment to her own words is a condition for the acceptance of the testimony on the part of the listener.

In scientific narrative, epistemic trust is composed of several actions based upon feelings of expectation, hope and confidence (Barbalet, 2011; Giddens, 1990; Rose-Ackerman, 2001), and also accuracy, sincerity, honesty and openness (Daukas, 2006; Williams, 2002). The agents could be the speaker and the listener, trust-giver and trust-taker, or simply partner and partner. These agents are always involved in cooperative activity and cooperative behavior. We take the testimony of others and we trust them to be a source of knowledge (Daukas, 2006). In Scanlon’s (2008) opinion, blame is essentially bound up with an impairment of relations with the offender and a withdrawal from the offender, which reflects that impairment. It involves the impairment of the relationship between subjects. The relationship is impaired because the offender’s intolerable practices make the person hard to trust.

In addition, the notion of trust does not necessarily imply the feeling of trust. As Hume (1738) points out, one may trust someone on the basis of pure knowledge: for example, I may trust a person to do me a favor, not because I feel that I can trust him or her, but because I know that he or she also needs a favor from me later on. Likewise, Hume gives his famous example of two rowers of a boat who pull the oars together without any explicit agreement or mutual trust, simply because they know that it is mutually advantageous to rely on one another.

Harris’ (2012) work shows how in the institution of the family, children trust parents because they consider them a reliable source of information and knowledge. But when children are not satisfied by the answer their parents give, they offer a counter-example or continue asking questions. Children in this institution use epistemic trust as a dialogical process with their parents because in this relationship between trust-giver and trust-taker, nobody is a passive receiver. It denotes a negotiation process based on this epistemic trust, where social functioning requires epistemic cooperation. This epistemic cooperation requires trust (Daukas, 2006).

For Bennett (2013) the reason for changing the terms of the relationship is out of prudence, for self-protection. When persons revise their expectations and intentions in a relationship, it is to change or modify the relationship. Friendship, for instance, is composed (completely or not) of several second-order emotions such as trust, confidence, closeness and sharing. Each of these second-order emotions regulates friendship, and second-order emotions probably act in different ranges for each friend. A relationship based on friendship is essentially composed of closeness and sharing, and is maintained thanks to trust in him or her (completely or in part). Each one of these aspects regulate and shape the relationship.

We cannot consider as a friend someone who does not treat us as a friend. This relationship is constituted by certain second-order emotions experience in relation to one another. When one of them changes her second-order emotion, the relationship changes. When, for some reason, we lose trust in our friend, this second-order emotion will probably be changed into a negative second-order emotion, like blame. This change of second-order emotion is to protect us from that person in the future. The change of this second-order emotion is useful for revising a relationship when someone acts in a way that makes it inadvisable to continue to have the same relationship with her (Bennett, 2013). In Bennett’s example, the blame expressed towards murderers and rapists takes the form of not trusting them, not helping them and not hoping things go well for them. The second-order emotion of blame is converted into an expression of disapproval for some bad actions. For this, such relationships require the participants to have certain sorts of mutual attitudes of concern on.

The point is not just how a person will act in the future but what has happened in the past, and what it indicates about the person’s attitude toward the relationship (Scanlon, 2008). That is why epistemic trust is a good indicator of joint action. Trusting signals a sort of social bond between two persons engaged in a common plan of action, such as walking together in the same direction, with the same rhythm, co-workers enjoying an interesting conversation during the lunch break. Walking together is not the purpose, the conversation is. It is a social activity for sharing knowledge and obtaining information. This ‘walking together’ represents a second-order purpose, the basis to cooperate. If it is impossible to share this basis, it is impossible to construct a solid relationship in social life.

Social relationships, like romantic relationships or friendship, take form thanks to joint actions, which require explicit collaborative intentions and can be conceptualized in terms of social interactions based on trust. In this interaction with other people, the subject generates a sense of involvement (Barbalet, 2011; Kemper, 1978) and these involvements are constructed thanks to trust. Take Scanlon’s example: “I might, (…), cease to value spending time with him in the way one does with a friend, and I might revise my intentions to confide in him and to encourage him to confide in me” (2008, pp.129-130). This example of trust takes place in the context of an ongoing relationship characterized by shared interaction. To trust someone is to hold modified attitudes of this kind toward her.

Losing trust in someone means not sharing this knowledge and not obtaining this information. In romantic relationships, when a partner loses trust in her partner, she does not have to access to this information, or probably she does not trust her words. To obtain this information, or to solve this lack of trust, she feels the desire to spy, eavesdrop, or read her messages, mail or her personal diary.

The social bond constructed by trust is particularly interesting to observe in an intercultural context. In China there is a social phenomenon called guanxi: “An intricate and pervasive relation network which Chinese cultivate energetically, subtly, and imaginatively” (Sennett, 2012, p.135). For instance, a Chinese person who has newly arrived in a foreign city feels free to call a cousin’s friend to help her to look for a job and somewhere to sleep. Guanxi means trust in social relations and in social networks, where “my word is my bond”. This informal cohesion is composed by a dialogic exchange, a negotiated process based on epistemic trust. The main purpose of the guanxi relationship is not to trust others, but to obtain information and knowledge to live in a new context. Epistemic trust is present in guanxi as a second-order emotion, which constructs an emotional tie between two persons that don’t know each other in a direct form, but through a third person (e.g., a cousin).Guanxi is a distributed form of trust in every member of a person’s network (e.g., a cousin’s friend).

In everyday life, we ‘use’ this second-order emotion of trust in a similar way to guanxi. Epistemic trust is a dialogic social experience embodied in everyday practices. Communication requires that we presume testimony to be generally reliable, like a sense of epistemic trust in the others (Coady, 1992). When a conversation starts, we presume it will be reliable and fruitful. Otherwise the conversation will barely start. Epistemic trust is a social epistemic second-order emotion because it depends on appropriate attitudes toward others, as well as toward oneself, as epistemic agents (Daukas, 2006). Trust is only given to, or received from, persons showing certain interpersonal and mutual dependence in particular situations. Responsibility emerges from a reciprocal understanding of the situation as a situation where the person involved is answering to the epistemic needs manifested by others.

Finally, the second-order emotion of trust is not just a dialogical exchange displayed in verbal communication, but also in nonverbal communication. We trust someone because of a knowing look or a positive bodily predisposition. Social gestures and gazes of understanding are small actions that allow us to internalize this second-order emotion in our relationships. Norms of trust are internalized in social practices. We do not think every time we meet the same person, “I trust her” or “I don’t trust her”, because we have internalized our trust or lack of trust for her in different episodes!

Final remarks

As social scientists, Jasper (2011) suggests that we need to build on people’s natural language labels, but we also need to make better analytic distinctions about them. In emotion research, we need to consider second-order emotions to construct a complete theory of emotions following the work of Georg Simmel (1964) and his notion of secondorder forms of sociation. The complexity of studying emotions comes from taking into account communication and language as a rich source of data. Second-order emotions are the structure of these emotional experiences in narratives. We agree with Harré (2000) that language is the primary instrument through which such creations and modifications come to be. The narratives of our lives are fundamentally perspectival, and they can profoundly distort reality, truth, objectivity, and our ideas of what it is to be a person (Goldie, 2012). Moreover, emotion is a work-in-progress, as we build emotion simultaneously with our use of language to refer to emotional states (Wilce, 2014).

The forms of sociation enable us to understand how individuals perceive society and express themselves in social terms. Some of these forms are ‘special’, or second-order forms, that represent analytical abstractions of patterns of relation. To exemplify this, Simmel introduces gratitude and faithfulness, while Cantó-Milà looks at gratitude in recent work (2012), and it is on this that we have based our second-order emotion of trust. We have observed how in recent years more researchers have focused on the second-order forms of emotions.

The central aspects of trust are emotional and epistemic. Trust as a second-order emotion constitutes an epistemic space for people to share knowledge in. Trust is not only needed to gain access to knowledge, but it is essential to becoming a socially situated self, to engaging in public conversations with others. Trust as a second-order emotion is internalized through social practices, in our discourses, and in our practices. In the model, we have observed how trust is essential to becoming a socially situated self. Without trust, the person feels negative emotions.

Although trust and distrust are fundamental to the construction of social ties, we still need to understand how they work or how to figure out their nuances and distinctions. Trust becomes a second-order emotion, which brings the future into the horizon of the expected. Trust contributes to narrowing the horizons of the expectable so that our social life becomes possible. The construction and maintenance of the emotion and bond of trust appears to be crucial in our daily relationships, and we dare to suggest that it seems like an important issue far beyond the system of intimacy. We hope that future research will use this model to analyze different contexts where trust is relevant.