Grief and the Unity of Emotion

The nature of grief presents a particular challenge to the view that emotions are unified psychological states.Grief can include all manner of potential ingredients, changes markedly over time, and has temporal gaps.In this paper, I focus exclusively on the relevant phenomenology and show how an experience of grief still amounts to a unified whole.I begin by endorsing the view that grief is a temporally extended process rather than an episode or state.I go on to argue that what unifies this process and singles it out as one of grieving is not -as has been suggested-its narrative structure.Rather, various aspects of grief all involve recognizing and responding to a wide-ranging, dynamic, and singular disturbance of life-possibilities, where recognition and response are inextricable from each other.This disturbance is 'held together' by relationships of non-propositional implication.I suggest that the same approach can be applied to emotions more generally.that emotions are judgments turns out to be trivially true.Scarantino (2010) therefore objects to what he calls the "elastic strategy": we stretch 'judgment' as much as we need to, until it fits around emotion.When stretched that much, it also encompasses the 'feelings' that others contrast with 'judgments'.Consequently, we fail to distinguish between importantly different aspects of emotion.If we instead acknowledge their distinctness and concede that emotions incorporate both, and perhaps a lot more besides, we are left with the problem I am concerned with here: how they together constitute a distinctive type of psychological state, rather than a mere bundle of other phenomena.Prinz (2004, p.18) thus describes the movement between a "problem of plenty" and a "problem of parts".The problem of plenty arises when we attempt to accommodate all the different ingredients of emotion, but in so doing fail to account for how they "hang together" and lose sight of the overall phenomenon.The problem of parts then surfaces when we try to tidy things up by identifying those ingredients that are essential.The most plausible answer seems to be 'all', taking us back to the problem of plenty. 1 In addition to this, there is the problem of specifying what it is for things to "hang together" in the right way (Dancy, 2014).The players in a football team, the two sides of a coin, the Earth and the Moon, the Morning Star and the Evening Star, and the numbers 1 and 2 all relate to each other in importantly different ways.Likewise, there are many ways in which the constituents of an emotion might be said to "hang together".So the task is not only to show that they do but also how they do.One response is to maintain that the two principal ingredients of emotion are in fact one and the same.Instead of contrasting bodily feelings with world-directed intentionality, one could argue that some or even all bodily feelings have world-directed intentionality (Goldie, 2000(Goldie, , 2002;;Ratcliffe, 2005Ratcliffe, , 2008)).Once this is admitted, specifically emotional judgments can be identified with emotional feelings, or at least with certain types of emotional feeling.A recent approach along these lines is proposed by Deonna and Teroni (2015).They maintain that different types of emotion consist of different types of evaluative attitude.Fearing and hoping are thus comparable to perceiving and remembering, insofar as they are sui generis types of mental state. 2 Next, they endorse the view that emotions involve modes of "action-readiness" (e.g.Frijda, 2007Frijda, /2013) ) and propose that a type of evaluative attitude is comprised of a distinctive set of diffuse bodily dispositions.Given that these dispositions are to some degree phenomenologically accessible, evaluative attitudes are inseparable from what could just as well be termed 'bodily feelings'.Suppose -for the sake of argument-that something like this is right (and I think it is).At this point, it looks as though we can avoid a game where player 1 says that emotions consist of p, player 2 says that they instead consist of q, player 3 proposes that they consist of both p and q, and player 4 then asks how, if at all, p and q stick together.Instead, we have an identity claim: p is q.Even if some or all emotions have additional ingredients, it could be argued that p/q is sufficient for being in a specifically emotional state, and also that a certain type of p/q is sufficient for having an emotion of a given type.This might well work for some emotions, perhaps even the majority.But, as I will show, it cannot accommodate grief.One option is to maintain that grief is different in kind from most of those psychological states/episodes that are labeled as 'emotions'.The problem can then be circumvented by conceding that grief is not an integrated episode, while continuing to insist that other emotions are.However, the unity of grief should not be dismissed prematurely and, in what follows, I will show how grief is, after all, an integrated experience.I will add that what applies to grief plausibly applies to emotions more