Sentiments are hard to define conceptually, although Gervais & Fessler (G&F) make valiant and reasonable attempts to do so. And given that a sentiment is variously described as a “syndrome” or a “network” or a “deep structure,” sentiments may be difficult to define operationally too. (It is not yet clear that sentiments are a readily measurable psychological construct.) In order for the sentiment concept to catch on again, I suspect that it will have to be defined more precisely and tethered more rigorously to a computational approach to motivational systems (Cosmides & Tooby Reference Cosmides and Tooby2013; Tooby et al. Reference Tooby, Cosmides, Sell, Lieberman, Sznycer and Elliot2008). Still, G&F's analysis of contempt is provocative; and it suggests that sentiments – whatever they are exactly – may offer a useful lens through which to examine human affect and human motivation.

Of course, if the sentiment concept is to be influential, it must be relevant to more than just the psychology of contempt. It must be applicable to a wider range of social relationships and motivational systems pertaining to those relationships. So let us consider carefully whether the sentiment concept might apply to something that is very different from contempt. Let us talk about love.

Echoing others (e.g., Shand Reference Shand1920), G&F identify love as a prototypic sentiment. This assertion seems superficially appealing, but it is probably not quite right. Love is perhaps too diffuse a construct to fit sensibly within an evolutionary analysis of the sort offered by G&F. Love comes in a variety of different flavors (e.g., romantic love, filial love, parental love) that are specific to functionally different kinds of relationships and that dispose individuals toward different kinds of behavioral responses (Shaver et al. Reference Shaver, Morgan and Wu1996). But even if the vague folk concept of love does not qualify as a sentiment, each relationship-specific form of love might make the cut. With that in mind, I focus on one specific form of love: parental love. How do the characteristic features of sentiments fit with what we know about the motivational psychology of parental care?

Sentiments are characterized as functionally specialized networks of attitudes and emotions that evolved in response to selection pressures arising within specific kinds of relationships. Does this apply to parental care? Yes. Parental caregiving responses are products of genetically coded neural mechanisms and neurochemical processes that are, to some extent, distinct from those associated with other motivational systems (Feldman Reference Feldman2016; Mileva-Seitz et al. Reference Mileva-Seitz, Bakermans-Kranenburg and van IJzendoorn2016; Rilling Reference Rilling2013). This underlying physiology appears to have evolved in response to the unique fitness implications associated with the provision of parental care to offspring (Kenrick et al. Reference Kenrick, Griskevicius, Neuberg and Schaller2010; Preston Reference Preston2013).

Sentiments are characterized as enduring, emotionally textured responses. Does this apply to parental care? Yes. There are stable individual differences in individuals' affective responses to children (Buckels et al. Reference Buckels, Beall, Hofer, Lin, Zhou and Schaller2015). Attitudes constitute part of this constellation of affective responses, but there is more to it than mere liking or disliking. The parental disposition is characterized also by a capacity to experience very particular, functionally specific emotional responses – such as tenderness, which is empirically distinct from other compassionate responses (Buckels et al. Reference Buckels, Beall, Hofer, Lin, Zhou and Schaller2015; Kalawski Reference Kalawski2010; Lishner et al. Reference Lishner, Batson and Huss2011).

Sentiments are characterized as being emotionally pluripotent, manifesting in different emotional expressions under different contextual circumstances. Does this apply to parental care? Yes. The perception of young children elicits a tenderness response, which is subjectively experienced as a rewarding emotional state (Buckels et al. Reference Buckels, Beall, Hofer, Lin, Zhou and Schaller2015; Kalawski Reference Kalawski2010), and may facilitate nurturing behaviors. But parental care is characterized not only by nurturing behaviors but by protective behaviors, too, which may manifest in risk-aversion and antagonistic responses to potentially threatening things (Eibach & Mock Reference Eibach and Mock2011; Fessler et al. Reference Fessler, Holbrook, Pollack and Hahn-Holbrook2014; Gilead & Lieberman Reference Gilead and Lieberman2014; Hahn-Holbrook et al. Reference Hahn-Holbrook, Holt-Lunstad, Holbrook, Coyne and Lawson2011). These protective responses are typically associated with entirely different kinds of emotions, such as fear and disgust and anger.

Sentiments are characterized as being responsive to functionally relevant relational cues. Does this apply to parental care? Yes; and here things get a bit more complicated. Parental responses – including tender responses to children and aversive responses to the broader environment – are triggered not just by the perception of cues indicating the presence of one's own offspring, but by the perception of human infants more generally, and even by things that merely mimic prototypic features of human infants, such as baby nonhuman animals or adults with baby-faced features (Buckels et al. Reference Buckels, Beall, Hofer, Lin, Zhou and Schaller2015; Glocker et al. Reference Glocker, Langleben, Ruparel, Loughead, Gur and Sachser2009; Sherman et al. Reference Sherman, Haidt and Coan2009). These responses are exhibited not just by parents, but by non-parents too.

In sum, there is an evolved “deep structure” of parental love that seems to fit with G&F's conceptualization of sentiments; but this parental sentiment is directed toward an unusually large and fuzzy category of relational objects. Indeed, one need not have had any prior interaction with – or even any meaningful knowledge of – an object in order for it to elicit a parental affective response. This contrasts with other alleged sentiments, such as contempt and hate and romantic love, which are typically directed toward specific individuals with whom one has had some prior interaction or at least some prior knowledge. So is parental love a sentiment? I am not sure. Might there be different kinds of sentiments – some that require input from prior experience with particular relational objects, and others that do not? Again, I am not sure. What I am sure of is this: Before the sentiment concept can be applied productively to a broad range of motivational systems and affective experiences, some rigorous conceptual work needs to be undertaken. G&F have taken some necessary and stimulating first steps, and I commend them for it. The hard work remains to be done.