Fragments on Empathy

Recent conversations with strong individuals have prompted me to post some fragments from a paper that I started writing last year but never completed, nor refined. The original structure of the paper was meant to a. lay out the problem, b. provide a historical context, c. map the historical context onto the contemporary debate, d. show how the relevant connections provide insight. I only made it through an abbreviated laying out of the problem and a rough historical overview. Comments would be most welcome:

The Problem
The fundamental issue at hand is the need to reconcile two strongly contrasting intuitions regarding social perception. The first is that an essential property of mental states is that they are not accessible. Prima facie, the way I know about your mental state is fundamentally different from the way I perceive medium sized objects. The second is that the mental states of others, in some basic sense, are directly perceivable, and that understanding the expressive behavior of others is neither a matter of theorizing about it nor sequential processes of simulation and projection. The first intuition evokes Cartesian conceptions of the mental, while the second echoes Wittgenstein’s anti-Cartesian challenge: “Do you look within yourself, in order to recognize the fury in his face?” (Wittgenstein 1980, sect. 927).
Phenomenologists are quick to invoke Wittgenstein’s blunt common sense approach, but their discussions of social perception can result in puzzling conceptual formulations. Gallagher and Zahavi, for instance, claim “that the expressive relation between mental phenomena and behavior is stronger than that of a mere contingent causal connection, though weaker than that of identity” (Gallagher & Zahavi 2008, 185). What could this mean? How are we to conceive of such a relation? I appreciate this attempt to balance the aforementioned contrasting intuitions, but in order for phenomenological considerations to be significant, further conceptual clarification is needed.
In order to provide the necessary clarification, first we need to identify the concepts that are causing the confusion. Clearly, the concept ‘expression’ is crucial to the debate, as it provides the link between observable behavior and private mental states. But if we begin to explore this concept, we find that our ability to discern our own mental states becomes caught up in any debate about how to discern the mental states of others. As Davidson has pointed out, “If the mental states of other are known only through their behavioral…manifestation, while this is not true of our own mental states, why should we think our own mental states are anything like those of others?” (Davidson 2001, 207). Thus, if phenomenology is dedicated to studying consciousness from the first person perspective, it may appear that it has nothing to contribute to an account of social perception.

The problem of social perception is nothing new in philosophy. Indeed, “the problem of other minds” is at least as old as Descartes. However, the contemporary debate between phenomenology and ‘analytic’ theories of mind can be made clearer by examining a very similar debate that was ongoing around the turn of the 20th century.
The concept of ‘expressivity’ did much of the heavy lifting in early explanations of how phenomena with both ‘inner’ and ‘outer’ reality are adequately grasped. At the turn of the century, philosophers, psychologists, and historians all saw empathy (einfühlung) and understanding (Verstehen) as intimately related concepts, both of which dealt with how we are able to grasp the ‘inner’ or ‘spiritual’ reality of something based on its ‘outer’ manifestation. Theodore Lipps, saw empathy as an especially important concept for explaining how aesthetic experience is constituted as meaningful. In Lipps (1903), our ability to understand and appreciate the beauty of external objects is directly linked to our ability to recognize other subjects as having minds. Lipps drew this connection “since for him the paradigm of beauty for human is the form, movement, and expressivity of the human body itself” (Steuber 2006, 7). This led Lipps to infer the existence of an underlying sub-personal mechanism that allows for the noninferential and quasi-perceptual character of aesthetic experience (ibid). Lipps characterized this sub-personal mechanism as ‘inner imitation’, and his primary intention was to causally explain why it is that we “move our legs when we see dancing” and “…” [citation needed].
While Lipps was interested in explaining the causal genesis of empathy through the function of basic sub-personal processes, other thinkers described empathy as a higher-level cognitive process. The early theorists in the hermeneutic tradition, namely Schleiermacher and Dilthey, also linked einfühlung and Verstehen, but had aims different from Lipps’. Schleiermacher saw empathy as the ability to understand other minds by imagining ourselves in various contexts [cite Scheleiermacher 1998, p. 92]. Dilthey expanded on this description, but opted for the term ‘reexperiencing’ (Nacherleben) to capture the methodological procedure of understanding others, guided by a variety of historical sources and by our knowledge of the general political, economic, social, and psychological state of affairs relevant for an adequate grasp of a particular historical time [cite Dilthey 1981, 172, 175-176]. As Steuber points out, “It was Dilthey’s…complete ‘psychologization’ of the interpretive process that made possible the common identification of the concept of understanding and empathy that we find at the beginning of the twentieth century” (2006, 11).
Stepping back for a moment, before examining Stein/Husserl’s phenomenological critique of the then prevailing views of empathy/understanding, we can see that at the turn of the century a similar twofold distinction within the concept of empathy was already taking shape. Taking Lipps as paradigmatic, one side of this distinction developed the concept of empathy as part of a psychological-causal-genetic account of what processes must take place within an individual when social perception occurs. The other side of this distinction, as exemplified by the hermeneutic tradition inaugurated by Schleiermacher and Dilthey, conceived of empathy as a highly cognitive process involved in interpreting the meaning of the signs, marks, and artifacts created by others. Lipps was interested understanding what sub-personal processes could cause social perception. Schleiermacher and Dilthey were interested in the personal processes that could allow for understanding others from disparate places and times. Before even considering what phenomenology might add to this characterization, it is noteworthy that contemporary empathy theorists have already mapped this framework onto the contemporary TT/ST debate. Steuber (2006) sees recent neurobiology as supporting Lipps’ conception of empathy, referring to this sub-personal, underlying causal mechanism as ‘basic empathy’—which he distinguishes from ‘reenactive empathy’, which allows us to “conceive of another person’s more complex social behavior as the behavior of a rational agent who acts for a reason,” which corresponds to the more robust concept of empathy developed in the hermeneutic tradition (20-21).
Edith Stein’s 1916 doctoral dissertation under Edmund Husserl, Zum Problem der Einfühlung (On The Problem of Empathy, trans. 1989) explores empathy as a phenomenon sui generis. On Stein’s account, Lipps’ theory of ‘inner imitation’ confuses social perception, or perception of ‘foreign experience’, with self-experience, or that which is ‘primordially given’. Her basic claim is that Lipps’ theory characterizes empathy as ‘convergence’ or ‘an experience of oneness’ between observer and what is observed. One way to understand this is by highlighting the difference between a memory and a flashback. When one has a memory, one perceives an experience that is no longer present. When one has a flashback, one experiences an experience as if it were literally present. In Stein’s terms, memory (along with fantasy) is an example of non-primordial experience—the subject has an experience of an experience that is not ‘fully given’ in the way that my present experience of my direct surroundings are given to me. A flashback, on the other hand, differs from a memory precisely because this line between non-primoridal and primordial experience is blurred. Empathy, in Stein’s sense, “can only be the non-primordial experience which announces a primordial one” (Stein 1989, 13-14). The potentially traumatic effects of a flashback are possible insofar as one is having an experience which is indistinguishable from the present, fully given experience typical of everyday perception. Thus, empathy shares the structure of memory and fantasy, whereas Lipps’ characterization of ‘inner imitation’ too easily leads to a conception of empathy having the same structure as flashback.
Lipps may also be understood as having confused the concepts of empathy and sympathy. Stein is careful to note that sympathy is a primordial experience whereas empathy is not, however this does not preclude the possibility of the two experiences coinciding. For example, there is a difference between my perceiving the grief in your facial expression and my feeling that grief. My empathetic perception of your grief may lead me to attribute a mental state to you, and it may serve as the basis for my theorizing about the larger context of your grief; but at no point is it necessary to this experience that I myself feel grief. If you are my friend or loved one, and I perceive that you are grieving, I myself, as someone who cares for your well-being, may also feel grief. But this primordial feeling of grief that I now have is better characterized as sympathy, or feeling with the other, and not empathy. Thus it is common for therapists to speak of the importance being empathetic while remaining unsympathetic. The job of the therapist is to understand the mental states of the other without being affected by them.
Lipps’ talk of ‘oneness’ and ‘convergence’ are especially problematic if carried out to its logical conclusion—telepathy. For if I truly was to ‘feel your pain’, the distinction between us as separate egos would be entirely erased. If I feel your pain as you feel it, then does it even make sense to call this empathy? The phenomenon we set out to explain, under the name ‘empathy’, was the phenomena whereby I am somehow aware of your mental states. If I am experiencing your mental states qua you and not qua I, this is not empathy.


  1. i love being referred to as a "strong individual"

    good paper. lots of powerful ideas therein. need to read over again a few times to make any fruitful comments.

  2. really just meant as a historical summary, first part of a first section of a paper