Empathy and Togetherness Online Compared to IRL: A Phenomenological Account

In this paper I aim to show with the aid of philosophers Edith Stein and Peter Goldie, how empathy and other social feelings are instantiated and developed in real life versus on the Internet. The examples of on-line communication show both how important the embodied aspects of empathy are and how empathy may be possible also in the cases of encountering personal stories rather than personal bodies. Since video meetings, social media, online gaming and other forms of interaction via digital tech-nologies are taking up an increasing part of our time, it is important to understand how such forms of social intercourse are different from in real life (IRL) meetings and why they can accordingly foster not only new communal bonds but also hatred and misunderstanding.


Introduction
In this paper, I would like to make a phenomenological contribution to the debates about the nature (and lack) of empathy and togetherness online by turning to theories developed by Edith Stein (2008Stein ( , 2010 and Peter Goldie (2000Goldie ( , 2011. Stein made her contributions to the phenomenology of empathy and sociality about a hundred years ago whereas Goldie's work on social emotions is more recent. Notwithstanding that their theories belong to different philosophical traditions and make use of different terminologies, I would like to point towards a basic predicament that they share: they both proceed from the assumption that feeling the presence of the other person is a necessary condition for having social experiences. Neither having empathy with the other person, nor having an experience together with the other person, can be achieved if we are not connected by way of feelings. My investigation and comparison of empathy and togetherness IRL versus online will be philosophical in nature. It rests on empirical studies and data taken from the literature I refer to, but I have not carried out any qualitative study of the behaviour or views of social media users or online gaming participants myself. Rather, the philosophical analysis is meant to serve as a possible starting point for conceptualizing empathy and togetherness in empirical studies, which could, in turn, serve to further sharpen or possibly change the phenomenological understanding of empathy and sociality in different environments. My hope is that the conceptual grid offered by way of the phenomenological theories of Stein and Goldie will be used in empirical studies exploring empathy and togetherness on the Internet to make us better understand digital communication in its many forms compared to IRL. Possibly, such a phenomenological theory of empathy could also be applied to make users of digital media aware of the pitfalls involved in digital encounters, similarly to how Magnus Englander has made use of phenomenological theories to design empathy training programs for caring professionals (Englander, 2019).

Empathy and Online Human Interaction
The testimonies of emotionally enflamed disputes and disrespectful language on social media sites are innumerable. Digital interfaces in the current set up seem not to facilitate empathic dialogue, at least not when the participants do not already agree on the matters commented upon and discussed. Agreeing and disagreeing online appear to be based on immediate emotional reactions that often prevent the parties from thinking clearly and treating each other with respect. There are many reasons for this, the most important one being probably that social media sites are not designed to foster mutual understanding and rational debates, but to provide maximum profit for the providing companies (Taplin, 2017). What catches our attention online are pictures and comments that we really, really like or really, really dislike, and Facebook, WhatsApp, WeChat, Instagram and others have made the most out of this basic aspect of human psychology (Eyal, 2014). The same is true for on-line gaming communities, another recent form of human interaction that is enjoyed by way of the technological interface of a mobile phone, computer or console screen (Alter, 2017). Since social media, gaming and other forms of interaction via digital technologies are taking up an increasing part of our time, it is important to understand how such forms of social intercourse are different from in real life (IRL) meetings and why they can accordingly foster not only new communal bonds but also hatred and misunderstanding. Having said this, it is also important to acknowledge that IRL meetings have their own problems and that empathy and genuine togetherness are not the only or even the most common forms of engaging with the other in the everyday world. Stress, anger, fear and selfishness make us disregard or treat other persons in morally problematic ways despite them being physically present in front of us. The face-to-face meeting does not guarantee mutual understanding and respectful interaction. Nevertheless, empathic understanding and the ways of caring for the other person that face-to-face encounters make possible form an apt starting point for studies of online interaction. Admittedly, empathy does not guarantee morally excellent behaviour -think of the sadist or the torturer using empathy to get to know the weak spots of their victims -but it is the most important gate of entrance to moral life that we have (Svenaeus, 2016).
Social understanding between persons take place on different levels of interaction, for example, it is common to distinguish between "low level" and "high level" empathy (Maibom, 2017). In the case of empathy, these two levels are often described as "affective" versus "cognitive" in character. This terminology is chosen to differentiate between automated affective responses and conscious attempts to understand the feelings of others. Whereas it makes sense to differentiate (at least) two levels of empathy and togetherness, it is unfortunate to present them as "affective" and "cognitive" for the simple reason that the first tends to be interpreted as "by way of feelings devoid of cognitive content" and the latter as "by way of thoughts in contrast to feelings". As I will try to show below, the levels of empathy should instead be understood as perceptual versus imaginative in character, distinguishing perceptual feelings from imaginative feelings by way of the two levels. The levels are connected in the sense that perceptual empathy is imaginative in a minimal sense whereas imaginative empathy is perceptually grounded, but it nevertheless makes sense to keep them analytically apart.
Whereas Stein's theory can be employed to understand the nature of low level empathy and togetherness, Goldie's contribution is necessary to fully explicate the nature of high level empathy and social interaction. This will 81 Empathy and Togetherness Online Compared to IRL Journal of Phenomenological Psychology 52 (2021) 78-95 become clear in the comparison of face-to-face meetings with digital interactions carried out below. As I will try to show, the digital interface not only makes it generally harder to achieve perceptual empathy, it also opens up avenues for certain forms of togetherness, based on group feelings and narrative imagination, which can become even more powerful and transformative than is the case IRL.

Empathy according to Stein
In my investigation of empathy and how it may break down when communication is technologically mediated, I will proceed from an empirically updated version of Edith Stein's phenomenological theory of empathy found in the book Zum Problem der Einfühlung published in 1917 (2008), but also by proceeding from complementary thoughts on bodily intentionality and intersubjectivity that we find in her Beiträge zur philosophischen Begründung der Psychologie und der Geisteswissenschaften published in 1922(2010. The significance of Stein's theory of empathy has recently been brought to attention (Magri and Moran, 2017;Svenaeus, 2016Svenaeus, , 2018Szanto and Moran, 2020), but many important aspects of her work regarding the nature and role of empathy and social interaction remain to be spelled out. In her work, Stein puts forward a theory of empathy, which is developed in the attempt to found a philosophical anthropology and moral philosophy. Her main sources of inspiration are Edmund Husserl (1976) and Max Scheler (2005), but the phenomenological empathy theory she ends up with is different from theirs and also from other influential theories in historical and contemporary debates on the subject (for an overview, see Maibom, 2017). Stein's suggestion for how to picture the phenomenon of empathy is based in the idea that empathy is a way of feeling oneself into the experiences of other persons (in German: "sich einfühlen"). Stein herself takes the feeling aspects of empathy to be very much granted by the terminology of "Einfühlung" and therefore does not much address the topic of how, exactly, the phenomenology of different types of feelings, which is found in the two works I have referred to above, is linked to the phenomenology of empathy she presents. Empathy, according to Stein, is an experience of another person's experience, and the experiences ("Erlebnisse") that she considers to belong to empathy are various forms of feelings. They are perceptual and imaginative feelings on the part of the empathizer and feelings that are expressed bodily on the part of the empathee (the person who is empathized with). Specifically, Stein takes empathy to be a three-step process in which the experience of the other person (the empathee) (1) emerges to the empathizer as an experience had by the empathee, the empathizer then (2) follows the experience of the empathee through, in order to (3) return to a more comprehensive understanding of the meaning of the experience in question (Stein, 2008: 18-19).
Stein reserves a special name for the fundamental form of empathy enacted by way of the lived body ("Leib"), namely "perceptually felt" empathy ("Empfindungseinfühlung") (2008: 74-80). The lived body is not the physical body studied from the third-person perspective of the scientists ("Körper"), but the body being experienced from the first-person perspective of its owner, making us present in the life world together with others (Merleau-Ponty, 2012). Perceptual empathy is a process of recognition and understanding that takes place on the level of embodied existence when one lived body feels and perceives the presence of another lived body and follows its experiences through in a spontaneous manner (this represents a second-person perspective on the body). This process relies on what Stein calls fields of sensation ("Empfindungsfelder") through which perceptual feelings protrude from the foreign lived body and draw me into its presence (2008: 74-75). When I see a hand resting on a table I see immediately that it is a part of a lived body that is more or less tense or relaxed and that presses itself against the table in contrast to the book that lies beside it (2008: 75). Perceptual empathy becomes even stronger in character when we watch foreign bodies move in a manner that is characteristic for human embodiment, and it is also facilitated by emotional expressions, particularly through facial expressions and gestures that humans show and perform. According to Stein, the empathy process of three successive perceptual steps is fundamental to enter into higher level empathy involving imaginative steps to understand the reasons why other persons express various feelings (Svenaeus, 2018). In the model of empathy I develop in this paper based on the phenomenology of Stein (and Goldie) I therefore distinguish between three steps in the empathy process taking place on two levels: perceptual and imaginative empathy (six steps altogether).
Stein considers already step one in the process of perceptual empathy to deserve the name of empathy, but this is perhaps debatable proceeding from a contemporary understanding of the concept (Stein, 2008: 19). Do I really empathize with another person simply by seeing or hearing that she is angry, sad, happy, embarrassed or in pain? If this is the case, all forms of perceiving and encountering other persons are empathic in the sense of involving step-one empathy. Whereas this was the understanding of the terminology as employed by the early phenomenologists (Zahavi, 2014), my interest concerns empathy through all the three steps that Stein identify, and on one or both of the two levels (perceptual and imaginative empathy).
The key to understanding the empathy process, according to Stein, is to show how the three steps of empathy are attuned (feeling based) in nature, since the affective qualities provide the energy and logic by way of which the empathy process is not only inaugurated but also proceeds through the steps and levels by way of what could be called an empathetic drive. We spontaneously feel along with other persons and imagine what it would be like to be in their position. The feeling alongside (in the footsteps of) the other in empathy is neither to be confused with feeling together in the sense of having a mutual experience of the same thing, nor with what Scheler calls "Einsfühlung" (feelingat-one-with-the-other) (2008: 25-29). It is also not to be confused with that which both Stein and Scheler names sympathy ("Sympathie" or "Mitfühlen"), which is the empathizer's feeling of compassion for the empathee (Stein, 2008: 24-25). Empathy is neither a matter of sharing the experiences (feelings) of the empathee in the strict sense, nor is it necessary a compassionate feeling for her. Sympathy, according to Stein, requires empathy through all three steps on the two levels to be in place, since it is a caring about the other person in and through getting to know her experiences and personality by way of empathy (possibly in combination with other ways of getting to know her, such as verbal communication) (Stein, 2008: 102-104, 116-130). In such cases, the empathy process is not only motivated by a care for the wellbeing of the other person (a common way of linking sympathy to empathy), but is also employed as a high road to getting to know and caring for a specific person.
The steps that Stein discerns in the empathy process are typically reiterated; step three will serve as a new step one on a higher level -the imaginative in contrast to the perceptual level. The steps could also, however, be supplemented by other ways of engaging with the other person, such as talking to her or starting to do something for/to her or together with her. These acts will transform empathy into communication-and/or action-based forms of togetherness that Stein explores more in detail in the second part of Beiträge on communal being (Stein, 2010: 110 ff.). However, even though Stein restricts the empathy experience to the model specified above -steps and levels that do not include conversation and coordinated actions between the parties -a form of tacit communication is arguably present already in the empathy process, provided the empathee recognizes that she is being empathized with and therefore directs her expressive behaviour towards the empathizer in the process. And the empathic feeling alongside the experiences of the other person will be at work also in many "empathy plus" forms of human interaction, which, in addition to feeling and understanding the experiences of the other person, also involve talking to her and acting together in the world. As I pointed out above, perceptual empathy, at least in the step-one sense, needs to be at work in such cooperative experiences, since the first person must feel and understand that the other person is doing the same thing he is doing, and, reciprocally, in order for the cooperation to be precisely a joint endeavour (Stein, 2010: 156, 162, 202; on sociality in Stein, see Szanto, 2015).
Many contemporary so called simulation-based accounts of empathy proceed from the idea that the empathizer imagines what it would be like to be in the empathee's shoes without the empathee necessarily being present to the empathizer when empathizing (Goldman, 2006;Stueber, 2006). I will return more in detail to such theories when presenting the account put forward by Peter Goldie that belongs to this group (2000,2011). In this way one could empathize not only with persons who are not physically present in the room, but possibly with the dead or with fictional characters in a book. In contrast to this, the examples of empathy that Stein discusses are all based on the perceptual emergence of the other person in front of the empathizer. According to Stein, this is because all forms of empathy are put in motion by lived bodily expressions addressing themselves to the empathizer (humans or other animals expressing feelings through their body language, see Stein, 2008: 76 ff.). Stein takes all cases of empathy to be basically perceptual in character, meaning they are bodily felt experiences of other living bodies. Such an empathy theory does not disqualify the role of literature or human imagination in gaining a deeper understanding of other persons, it only denies that what I am doing when I am imagining what my friend would think about this book that I am reading, or, what a character in the book would think about my friend, are cases of empathy in themselves. As we will see in what follows, Goldie takes a different position regarding this matter and current developments in online communities could be interpreted to strengthen a view of empathy as based on narrative imagination also in cases when lived bodily interaction is restricted or absent.
Stein does not consider the examples of painted pictures or photographs displaying living bodies, although such examples originally served to illustrate the process of Einfühlung in the German tradition she is working in (for instance in the empathy theory of Theodor Lipps, see Coplan and Goldie, 2011). Even more pressing examples of technologically-mediated empathy are contemporary ways of encountering other persons by way of telephone or Skype conversations, or in social media encounters, which I will return to in a moment. In my view, Stein's theory is open to interpret telephone or Skype conversations in terms of empathy since they include mediations of the lived, expressive human body that is seen and/or heard in the encounter. However, that such ways of communicating include at least basic possibilities of empathy does not mean that they are as robust and fine-tuned as face-to-face IRL meetings are. Empathy according to Goldie Emotions are normally defined in contemporary philosophical literature as feelings that carry cognitive content; that is, the feelings we have that are about things in the world. If I love or hate a person this emotion expresses certain beliefs about how the world is constituted and also about how it ought to be (what I care about in the world). In contrast to the emotions some other feelings appear to lack cognitive content. If I am feeling pain this pain is not about anything in the world, although the feeling might be said to contain a certain evaluation of my present state as being bad and to be changed. The distinction between feelings that are emotions and feelings that are not emotions is not always clear-cut. I can be tired in the sense of being sleepy (not an emotion) but I can also be tired about something in the world (tired of my job, for instance). To some extent such fuzzy borders may depend upon us using the same word in describing different types of feelings, but the fuzziness may also be taken as an indication that emotions may carry cognitive content in different degrees that is made more or less explicit by the feeling in question. This is an important lesson to integrate into theories about empathy, a lesson that is acknowledged and elucidated in the model articulated by Stein.
Goldie's perspective in The Emotions: A Philosophical Exploration is, to my mind, phenomenological, even though he does not refer much to the philosophers of the phenomenological tradition (2000). In a phenomenological spirit, he opens his investigation by stating that in order to understand what emotions are we need to adopt "a personal perspective or point of view" (Goldie, 1990: 1). This first-person perspective, as Goldie explains, is not subjective in contrast to the perspective of the other person; instead, it presupposes a multi-personal point of view in which I occupy one unique point of view on a common and shared world. When I see you blush -appearing as another second personmy first thought will be that you are ashamed of something -and I may well already know what this is, since we share a situation in the world together. It is also possible for me to adopt a third-person scientific perspective on your appearance in which I explore the causal origins of your cheeks turning redincreased blood flow caused by neural events -but this is not the way other persons originally appear in the world to me. This distinction is important in discussing empathy, and the phenomenologist will insist that the first-person perspective precedes the third-person perspective in encounters with other persons (Zahavi 2014). In other words, I do not first consider the appearance of the other person in a scientific way to reach a conclusion about her state -that she is minded and has certain feelings and thoughts -the other appears as another person from the very start in my perception of her (Zahavi, 2011). The second-person perspective is certainly different from the first-person perspective; only I can feel and know my perspective as my own, but this is a difference within the personal point of view, argues the phenomenologist (and Goldie).
Goldie views emotions as bodily experiences in the sense that it always feels a certain way to be, for example, sad or happy. Emotions have a bodily dimension and they change my physiology in a way that is also, at least in the standard case, visible to others by way of my bodily expressions (facial expression, bodily posture and movements). When I undergo an emotional experience I feel it in my body, although in many cases I am not focused on the bodily sensations in question. This is no coincidence, since the emotional experience is directed towards something in the world that is made present to me in a vivid and attention-consuming way through the feeling in question -the thing I am happy or sad about. Goldie talks about emotions as ways of "feeling towards" states of the world in a kind of "imaginative perception", which he characterizes as a "thinking of with bodily feeling" (Goldie, 2000, chapter 2 and 3). Emotions are intentional in the phenomenological sense, they present a state of the world to me, and they are also normative, since they include evaluations of how I feel about this judgement.
A key issue here, which Goldie does not explore in any detail in his book, is how the bodily dimension contributes to the intentionality of the emotions (Ratcliffe, 2008). Summing up some aspects of the phenomenology of affective intentionality found in the first part of Stein's Beiträge zur philosophischen Begründung der Psychologie und der Geisteswissenschaften (2010), I would say that the bodily element of emotions does so by way of 1. Focusing our attention on what is valuable and relevant in the environment for us; 2. Resonating with bodily-expressive patterns that we perceive in other persons (or in other creatures or features of our environment that resemble human bodies); and 3. Responding to features of our environment, including the expressions of other persons, which should make us take action. These elements are the corner stones of what I would like to call "affective schemas", which provide the basic drive and direction for the emotions when they take on meaning as being about the world in "feeling towards" different things, as Goldie puts it (Svenaeus 2018).
Affective schemas are an important source of the normative element in emotions, since they provide a basic evaluation of how we find the state of the world to be (good or bad) and what needs to be done in a situation to change the predicament of ourselves and/or others. The work performed by affective schemas are influenced by many things pertaining to the history and context of the situation in which they are at work -the personal history and developed characteristics of the person having the emotion, for instance -and they do 87 Empathy and Togetherness Online Compared to IRL Journal of Phenomenological Psychology 52 (2021) 78-95 not predict in detail how a particular emotion had by a particular person in a particular situation will develop. Nevertheless, the schemas provide typical ways in which we feel and judge key situations in life that demand our attention. Affective schemas are likely present from birth in rudimentary forms that are then diversified and made immensely more complex during the first years of human life (Colombetti and Thompson, 2005;Iacoboni 2008). Affective schemas are related to what in the phenomenological tradition is known as "body schemas": systems of sensory-motor processes that constantly regulate our posture and movements, processes that function without reflective awareness which make the world appear to us as bearing significance in movement and action (Colombetti, 2014;Gallagher, 2005).
Goldie does not regard empathy as an emotion per se, since it is not typically an enduring state of a person's (embodied) mind in the same way that being sad, happy or angry about a certain condition in the world typically is. He does, however, regard sympathy as an emotion, since the latter consists in a longer lasting disposition to care about the wellbeing of another person. I can have sympathy for another person without feeling anything about her at the moment, but the sympathy makes me disposed to have episodes of sympathetic feelings towards the person and also to act on his behalf should certain situations arise (Goldie, 2000: 213 ff.). Goldie is right that empathy is not a disposition to feel, but rather an experience of feeling towards the other person that is carried out during a limited time span (a feeling episode). Goldie defines empathy as "a process or procedure by which a person centrally imagines the narrative (the thoughts, feeling and emotions) of another person" (Goldie, 2000: 195). This imagining, I would argue, perhaps in contrast to Goldie, is an emotional experience: an experience of feeling towards the other person in a certain way, "thinking about her with feeling", to use Goldie's own expressions (Goldie, 2000: 19).
Goldie's characterization of empathy needs phenomenological support (this is what I am trying to do with the help of Stein), since it disregards the perceptual feeling towards the other person in the face-to-face encounter that is, if not necessary, then at least typical and in many ways essential for the empathic process. His theory also presents the task of the empathizer in a way that is perhaps too demanding. Empathy, according to Goldie, is not only about imagining being in the other person's shoes (what it would be like for me to be in his position), it is about imagining what it would be like being her, and this may, actually, turn out to be an impossible project if we are not very similar to the persons we are to empathize with or, alternatively, if the situation and possibilities we are to consider are not very straightforward and noncomplex (see Goldie's modification of his previous account in Goldie, 2011).

Empathy and Communal Feelings in Digital Life
As I have tried to show with the help of Stein and Goldie, empathy rests on a spontaneous bodily-affective experience in face-to-face meetings between persons.
Step one empathy is basic to all forms of human interaction and communication, whereas empathy in the more robust sense of feeling alongside and understanding what the other person is experiencing, rests on perceptual empathy in all three steps.
Step one perceptual empathy happens more or less spontaneously according to affective schemas, whereas the second step of feeling alongside the emapthee is enhanced or possibly blocked by way of and according to the empathizer's will and emotional situation. It is harder to empathize with persons that one is scarred of, or do not like, and it is harder to empathize when one is busy with other things that consume one's emotional attention. On top of perceptual empathy, the empathizer may enter into more cognitively informed ways of feeling alongside the empathee by way of imagination. Such higher-level cases of empathy I have called imaginative empathy, since they rest on feelings that have cognitive content (emotions). Perceptual empathy is also imaginative in the minimal sense of going along with the feeling of the empathee in the second step, but this does not yet include any imaginative thoughts about why she is feeling so or so. Perceptual empathy spontaneously develops into imaginative empathy and the two forms may therefore be hard to tell apart in the individual cases (Svenaeus, 206, 2018). The explication of the different steps and stages of the empathy process, and other forms of feeling-based understanding of the other person, can help us see why empathy and dialogue face difficulties on the Internet. The technological interface does not provide the necessary bodily basis for perceptual empathy to occur, and in the interactions between persons that despite this do occur, there is often a lack of mutual understanding and respect for this very reason (Turkle, 2011;Twenge, 2017;Vaidhyanathan, 2018). The other person is not present in flesh in social media or gaming communities and this is probably why the comments and reactions found there often display a lack of empathy. It is much easier to write something nasty to or about another person when one does not have to witness her emotional reaction to the comment. This is especially so if one does not have prior IRL experiences of spending time with the person in question and consequently has not developed care or respect for her.
In a recent article, Lucy Osler argues, with the aid of one of Stein's phenomenological compatriots, Gerda Walther, that true fellow feeling and togetherness online are possible even though the IRL bodily presuppositions are absent (Osler, 2020;  the kind of phenomenological analysis of empathy and shared understanding that I have developed so far is twofold. First, she argues, on-line interaction is embodied even though in a more restricted and possibly different way than IRL meetings. Second, she argues, togetherness online is facilitated by other means than we-experiences IRL, and in power of the new technological opportunities it can actually lead to stronger emotional bonds in larger groups of participants. As should have become obvious already in my analysis of Skype conversations above, I fully agree with Osler's first point. We do not cease to be embodied beings when we position ourselves in front of a screen, and if the transmission of picture and sound through camera and microphone work well, with minimal delay and distortion, video meetings could get close to the real thing, although the sense and smell aspects are missing. However, Osler's two following examples of what she claims to be empathic online interactionavatars and chat apps -reveal how thin and restricted togetherness may become in cyberspace. Avatars, used in multiplayer on-line games, which you watch on the screen as you move and take action from your own first-person perspective, do not even come close to the experiences you will have if you, say, play hide and seek in a wood together with friends. Avatars in popular games played on PC s, consoles or mobile phones often resemble human bodies, but they do not express feelings in the same way that live human (or animal) bodies do. This is so regardless if the game is based on a first-person perspective (shooter games) or a third-person perspective (strategy games) or a combination of the two. As regards the possibilities for empathy to occur, it is also important to point out that a majority of the games being played today are based on some type of combat scenario that does not invite or encourage the kind of experiences that are necessary for feeling with the other person (neither empathy nor sympathy).
Osler's third example, chat apps in which the participants write messages to each other in instant dialogue or group communication, does obviously not have the necessary constituents for perceptual empathy, such as Stein describes it, to take place. What appears on the screen of the mobile phone or computer are words and possibly emoji, and even though these may very well give rise to strong feelings, the feelings are not the results of experiencing the lived expressive body of the other person. That other persons do not appear in bodily presence through chat apps is probably also the reason why it is easier to write mean and insulting things about them, since you do not have to experience their reactions in bodily presence. This is even more so on social media platforms where you comment on the contributions of others without entering into real time interactions, such as on Facebook, Instagram or various forms of blogs.
Despite the lack of lived-body based empathy in such encounters, Osler has an important point when she claims that a strong sense of togetherness is present in contemporary on-line communities. This togetherness rests on sharing a mutual interest or cause and is often formed by identifying a common enemy in contrast to the supportive feelings shared in the in-group. The MeToo-movement is an obvious example. The way empathy and sympathy is developed in such on line communities do not rest on bodily interaction but rather on the sharing of stories. Proceeding exclusively from Stein's theory we would not be in the position of naming such on-line interactions and fellow feelings empathic (and also not compassionate, since all sympathy rests on empathy according to Stein), but with the help of Goldie we are able to see that they may belong to the level that I have named imaginative empathy. As pointed out above, it is common to talk about empathy on different levels and even though perceptual empathy is standardly needed to found the process, it also appears possible in some cases to feel alongside an empathee even though she is not physically present to the empathizer. The vehicles for entering directly into imaginative empathy, in contrast to starting out with perceptual empathy, in such cases are not expressive bodies but rather touching stories. On-line communities are formed by the sharing of such stories and the empathic bonds that are formed through this sharing could in some cases become very strong, as witnessed by participants (Osler, 2020).
Imaginative empathy is a type of empathy that many simulation theoristssuch as Goldie -find paradigmatic. Empathy, according to such theories, is formed by imagining being in the empathee's position or perhaps even by imagining to be him -if this is really possible is another thing (Goldie, 2011). The way you can imagine this is by telling yourself, or reading or being told, a story, by which you come close to the experienced perspective of the author, or to the perspective of one of the protagonist appearing in the story. Such stories were probably told even before human beings developed that which we refer to as a culture, which predates the invention of writing. Stories seem to come naturally to us when we ponder upon the meaning of life and our destiny. We are narrative creatures from the very beginning and this has not changed with our newly developed digital habits.
Emotional stories -stories that give rise to strong feelings in readers -in different formats provide the most important way of getting attention on the Internet today alongside with emotionally loaded pictures. Such stories do not only give rise to empathic and compassionate feelings in readers, but also to hate, anger, disgust and rage being directed towards the author or towards people who figure in the stories. If a story is rich enough to express the emotions of its author in a way that displays her personality, or the personality of some protagonist in the story, in a nuanced and coherent way, such cases of imaginative empathy are arguably as vivid as embodied engagements. The question is if the current format of social media is supporting such attempts or if it is rather giving priority to stories that provoke strong feelings regardless of being true to the persons and events that are presented. In such cases the emotions that develop in the readers will rather be about types of persons, persons belonging to a certain group that the reader is a member of or that she is distancing herself from, than about actual persons with all the complexities and inconsistencies belonging to the real world. What gets lost with such a development is the individual voice, which expresses the bodily presence of a person and therefore evokes the perceptual empathy of the listener (even though she is reading rather than physically listening to the story).
A major problem with the internet as a forum for social bonding and understanding it that it makes it possible to encounter only the types of persons that belong to one's ingroup(s) and avoid all encounters with persons that are out. Perceptual empathy may also be hampered by segregation of living and working environments IRL, but the physical set up makes it more likely to encounter persons that do not belong to the same group and harbour views and emotional expressions that are different than one's own. In such encounters imaginative empathy is made robust and unpredictable when you, for instance, start talking to a stranger on the bus or in the super market. Working places, bars and parties make it easier to meet a partner in less strategic and ingroup-bound ways than Tinder and other dating apps do (Aho 2017). In this way meeting people IRL is both more real -built on lived bodily presenceand unpredictable in comparison with a more strategically selected ways of typifying one's social life online.
The point could be raised that people have enhanced their social understanding of other human beings by way of reading biographies, historical and fictional stories for hundreds of years. This is true, but it also needs to be taken into account that novels are typically richer and more complex in nature than social-media stories are. They allow for more complex personalities and ambivalent stories to be developed in which the reader may empathize and identify with more than one person and in ways that are less single minded. Social media and gaming will not relieve the participants of spending a considerable amount of time IRL, at least not yet, but there is every sign of the activities replacing the reading of all form of longer texts, not only of novels but also of newspapers articles and science books (Twenge 2017 to less perceptually and imaginary nuanced way of understanding other persons, which is a real problem for our current and future attempts to build empathic communities.

Conclusion
I have tried to show with the aid of Edith Stein and Peter Goldie, how empathy and other feelings about and with other persons are instantiated and developed in real life and on the Internet. The examples of on-line communication put the finger both on the basic embodied aspects of empathy and on the way empathy may be possible also in the cases of encountering personal stories rather than personal bodies. Empathy takes place in three steps on two interrelated levels: perceptual and imaginative empathy. Since empathy is triggered by bodily expressed feelings of the empathee, that are perceptually felt by the empathizer, it is an emotional process with an inherent drive to develop through the successive steps and stages. In the first step, the empathizer bodily perceives by way of an emotion that the empathee expresses a certain feeling (being in pain, scarred, happy, or whatever). In the second step, the empathizer feels with (alongside) the empathee by following her feeling through in order to return to a richer emotional understanding of the empathee's experience in a consecutive third step. The three steps of perceptual empathy are guided by what I have referred to as affective schemas, which provide patterns for how social feelings evolve and develop more or less preconsciously in meetings between expressive bodies. The third step of perceptual empathy is also the first step of imaginative empathy, by which the feelings explored by perceptual empathy may gain a much richer content taking into consideration what the experiences of the empathee are possibly about and how they are related to her personality. This is done by way of emotionally imagining being in the other person's place, or even being her, a process which is spurred and developed by way of narratives. The explication of the different stages of the empathy process, and other forms of feeling-based understanding of the other person, may help us see why empathy and dialogue face difficulties on the Internet. The technological interface does not always provide the necessary bodily basis for perceptual empathy to occur, and in the interactions between persons that despite this do occur, there is often a lack of mutual understanding and respect for this reason. Digital meetings by way of applications such as Skype clearly open up for perceptual (and imaginative) empathy to occur, since they include mediations of the lived, expressive human body that is seen and/or heard in the encounter. However, that such ways of communicating include basic possibilities of perceptual empathy does not mean that they are as robust and fine-tuned as face-to-face IRL meetings are.
In contrast to video meetings, the other person is normally not present by way of her lived expressive body in social media or gaming communities and this is probably also why the comments and reactions found there may display a lack of empathy. On the other hand, stories cultivated in online communities sometimes harbour possibilities for imaginative empathy also in cases in which the bodily-perceptual basis is minimal. Such stories may form strong communal bonds and sympathy for other persons. Nevertheless, the story telling in current digital set ups run the risk of degenerating to socially formed stereotypes to be empathized with, instead of encountering and empathizing with real other persons whose voices may resonate even though they are only present by words appearing on the screen.