In Emotion, Imagination, and the Limits of Reason, Talia Morag maintains that philosophers have characteristically over-intellectualized emotions by assuming that ‘the causal formation of emotions … involves some kind of evaluation of the worldly circumstances’ (265). Over the course of her rich and ambitious book, Morag criticizes this assumption and offers an alternative.Footnote ¹ For her, emotions involve a more primitive mental phenomenon, called the ‘associative imagination’, which she extracts from the work of Freud. On this alternative, emotions are constituted by a kind of transference, where the person unconsciously applies idiosyncratic ways of seeing earlier emotion-eliciting situations onto the present. Emotions are thus less rational and world-directed than philosophers have supposed. Overall, the book makes an informed and compelling case for the relevance of a host of psychoanalytic insights to the philosophy of emotion.

According to Morag, extant theories, including classical and modified judgmentalism, as well as sub-personal theories, fail to acknowledge how emotions are not always sensitive to reasoning and judgment. For example, despite forming the judgment that some remark does not makes sadness appropriate, a person may remain sad. Such examples illustrate ‘recalcitrance’ (38). However, a different person, or the same individual on a different occasion, may find her sadness to be lessened or eliminated by the same reasoning. Since these theories fail to recognize the qualified reason-sensitivity of emotion, they also face an explanatory burden: they fail to explain ‘why … this specific person is having this specific emotional episode’ (55). Morag calls this ‘the singularity question’.

The singularity question is indeed fundamental, and Morag successfully exposes the failure of many theories to provide an answer. Nevertheless, issues arise for her critique. Firstly, important theories have been overlooked. Morag does not consider the claim, defended by James, Stocker, Goldie and others, that emotions are constituted by sui generis states of feeling.Footnote ² This omission is puzzling, because feeling theorists emphasize that judgment is insufficient for emotion and that emotions resist alteration by reason and judgment.Footnote ³ Even more puzzling is the omission of Wollheim's work on emotion, which is a watershed in philosophizing about emotions using psychoanalytic concepts.Footnote ⁴ Secondly, Morag denies that the singularity question can be answered by adding desire, attachment, love, or concern (which she calls ‘care-factors’) to extant theories (33). Care-factors are claimed to be either equivalent to emotions, in which case appealing to them begs the singularity question, or equivalent to evaluations, in which case appealing to them is not sufficient explanation for why the emotion occurs (37, 52). These alternatives are not exhaustive, however, because some concepts of desire fall into neither category. Desires may be defined instead by a functional role, such as tending in certain conditions to cause specific sorts of efficacious behavior, and thereby, bringing about specific sorts of changes in the world.Footnote ⁵ On this view, desires need not have conceptual or evaluative content at all. While Morag may ultimately have responses to these overlooked approaches, the success of the book's negative argument depends on the success of those responses.

Morag's answer to the singularity question appeals to the intriguing concept of ‘associative imagination’, which is the capacity to connect words, images, and sensory data (sounds, smells, etc.) from memory with other words, images, and sensory data either in the present or in memory. The imagination finds connections among ‘fragments’, which are ‘parts of objects that the imagination considers in isolation from their surroundings’ (156). According to Morag, people constantly find associations despite being unaware of doing so. For example, a student has eyes that gleam like my friend's and a haircut like him too. When enough connections have ‘accumulated’, I immediately see the student's face in terms of my friend's (157). Adapting a term from Freud, Morag calls this phenomenon ‘condensation’. The similarity will strike me immediately without any need for inference or judgment. Since people are constantly seeing and remembering such connections, each person has ‘an imaginative or associative network of fragmented images or words, that is in constant flux’ (174).

The associative imagination is argued to provide causal, rather than rational, explanations of emotions. Morag postulates that infants begin their lives with a fixed repertoire of so-called ‘basic emotions’, such as anger, fear, disgust, etc. (202). She does not say enough about basic emotions, but it is clear that they involve physiological arousal and action tendencies with an evolutionary provenance. Later emotions occur when, due to the accumulation and condensation of associations, the person sees some present situation in terms of situations that triggered previous infantile basic emotions or other later emotions. The specific patterns of physiological arousal and action tendencies from the basic emotion are then transferred or ‘displaced’ upon the present situation (183). Morag says ‘if the latent emotion becomes heavily connected with fragments of the here-and-now situation, if enough such associations accumulate, the emotion is enlivened’ (175). For example, a person may be terrified in the presence of a police officer. Perhaps his voice or posture connects with previous fear-elicitors, such as angry parents. In this way, the singularity question can be answered. The specific associations that a person connects to the present situation causally explain which emotions she has in the present.

This account is meant to avoid the over-intellectualizing assumption that emotions necessarily involve evaluations while also recognizing the qualified reason-sensitivity of emotions. A person's associative paths frequently lead to certain emotions, which are ‘enlivened frequently and function as the terms in which situations are seen as emotionally salient’ (189). Morag calls these ‘nodal emotions’. Nodal emotions correspond to common themes in one's emotional life, like dangers, mothers, authority figures, public embarrassments, etc. Crucially, recognition of the theme is not part of the emotion itself, but is only discovered after the fact through the emotion's verbal expression. On its own, finding associative connections is simply a ‘blind operation… of the mind’, which does not presuppose rule-following, concept possession, or rational inference (178).Footnote ⁶ Thus, each emotion sees its object in terms of some oft-traversed node within the person's associative network, rather than a rational category of evaluation. Nevertheless, such emotions can yield to reasoning. Sometimes a rational judgment that an emotion is unwarranted or unfitting will be enough to ‘unplug’ the node from the present situation because it distracts the person from the associative connection (212). At other times, the connection remains salient and thus the emotion persists.

While there is undoubtedly truth in the claim that the associative imagination causally explains emotions, this claim also raises a question: what kind of explanation has been thereby provided? According to Morag, the associative imagination furnishes only ‘singular causation’, which describes some conditions C without which event E would not have happened, rather than general causal relations that describe how, given C, everything else equal, E will follow (12). Whereas general causal relations allow for rational prediction, a singular cause can only be assigned after its effect has already occurred. Morag's reason for endorsing singular causation is that the associative imagination (and hence emotional life) is unpredictable: ‘how exactly the imagination associates in every given set of circumstances is a contingent matter. The imagination can associate in new and unpredictable ways…’ (219). Yet, this premise is questionable. Associations could be inherently unpredictable because they are indeterministic, or could merely appear unpredictable because they are determined in accordance with general causal relations by a complex array of causal antecedents, including highly individual, subjective, or contextual ones.Footnote ⁷ These causal antecedents may nevertheless be difficult to recognize or measure. Why accept the former story over the latter? Morag argues that one's emotions can surprise others and that emotional patterns could change (73, 89). Neither point supports the inherent unpredictability of the associative imagination. Those who are surprised may fail to grasp the causal complexity behind an emotion. Further, general causal relations do not exclude the possibility of emotional change because the particular causal antecedents themselves could change.

Far from being tangential, these claims about predictability and causation are crucial to the book's overall argument. Morag alleges that the fact that the associative imagination can answer the singularity question supports her account uniquely within the landscape of the philosophy of emotion. However, if the associative imagination is inherently unpredictable, then it is a brute, arbitrary causal process. It is brute insofar it is cannot be subsumed under any causal process occuring at a more fundamental explanatory level. It is arbitrary insofar as there is no possible story at the same explanatory level of why the imagination associates in one way rather than another. In effect, Morag has answered the singularity question by so many determinate ways of saying ‘That is how she associated. End of story.’ No further answer can be given as to why a person associated as she did, and this non-answer is supposed to meet in principle any further demands for explanation. It is reasonable to reject as unhelpful and uninformative a causal process whose operations are so insulated from the rest of nature. In short, if the associative imagination is inherently unpredictable, the singularity question has not been answered satisfactorily.

Appealing to concepts of stability within an associative network will not eliminate the problem of arbitrariness. To begin, Morag asserts that themes within one's emotional life, which are due to the frequency of certain associative paths, can shed light on the person's particular ‘issues’ and ‘sensitivities’ (248). But themes cannot answer why the person associates as she does, because they do not ‘pre-exist in the network and do not play a causal role’ (199). Morag elsewhere invokes ‘habits’ of association (268). While habits would be non-arbitrary explanations for why a person associates as she does, it is not clear that appealing to habit is available here. A plausible way to define habit is as the reliable manifestation of some response in circumstances with recurrent features.Footnote ⁸ On this definition, habitual responses fall under general causal relations, and hence are predictable, even if the manifesting circumstances and their features are difficult to recognize or measure. So, unless Morag abandons the claim that the associative imagination is inherently unpredictable or provides a definition of habit that is consistent with this unpredictability, she loses the advantage that is supposed to support her view uniquely.

It is more credible that which associations one finds, and which condense in a particular situation, are determined by a complex array of subjective, environmental, and other factors. Freud himself offers one variety of causal determination, which is surprisingly overlooked in Morag's book. Freud sometimes invokes ‘wish’ to explain spontaneous thoughts, memories, and visualizations.Footnote ⁹ Wishes, which fall under Morag's category of care-factors, are neither emotions nor evaluations. They are states defined by the following functional role: rather than leading a person to adopt efficacious means towards their satisfaction, wishes cause him to imagine or visualize their satisfaction.Footnote ¹⁰ On this view, wishes lead the person to connect occurrent thoughts, images, sounds, etc., to those that have reliably created some experience of satisfaction or gratification in the past. If Freud is correct here, the associative imagination in some respects operates in accordance with general causal relations. Further, Freud can say more about why a person has certain wishes, because wishes originate from an organism's basic biological needs.Footnote ¹¹ While Morag rejects this appeal to ‘biological-cum-metaphysical drives’ to preserve the idea that the unconscious is undetermined by ‘any entity that is somehow separate or different from the person in question’, readers are not provided any rationale for this rejection or for the desideratum that the rejection supposedly meets (271).

Nevertheless, Emotion, Imagination, and the Limits of Reason remains a fascinating study of recalcitrant and other emotions. Morag's concept of associative imagination is a distinctive and promising explanation for why people have the emotions that they do. While the book's main argument is incomplete at points and fails to secure its conclusions against important and relevant doubts, it offers a broadly convincing picture of how intimately emotions are tied to the psychoanalytic unconscious. Those who are suspicious of, or persuaded by, philosophy's recent emphasis on the rationality of emotions will find in Morag's book a well-developed and fruitful alternative.Footnote ¹²