Natural Kinds, Social Constructions, and Ordinary Language

2.1 Objective Kinds

An objective kind is a class of things that have the possibility of being independent of human conceptualizations.^[7] It is this possible independence from human conceptualizations that defines an objective kind and its members – as something that is said to be “natural,” and therefore “real,” in contrast with something that is said to be “artificial,” and therefore something “unreal” or “unnatural” – rather than the fact that some sort of material or physical existence is predicated of them. In order for a class of things to have the possibility of being independent of human conceptualizations, and thus be an objective kind, it is necessary for members of that class to be unified in accordance with a feature of the world that has the possibility of being independent of human conceptualizations.

One possible unifying feature is that members of the same class all share some non-arbitrary physical property or properties, including causal properties of efficacy. Such properties unify the members of an objective kind as members of that kind, and make it possible for objective kinds to be what they are without the need of human conceptualizations. For example, Water is an objective kind. All things that are Water are thought to be real and to have non-arbitrary physical properties – being constituted by H₂O molecules – that identify them as members of the objective kind Water. Something that is not an objective kind is Chair. Although chairs have physical properties, these properties are arbitrary and do not contribute to identifying any chair as a Chair. What identify a chair as a Chair are not its arbitrary physical properties but rather its structural form and functional properties; but this does not preclude the fact that objective kinds have functional properties. For example, a vertebrate-heart is an objective kind. Not only do all vertebrate-hearts have physical properties that partially identify them as a Vertebrate-heart, they also all share the function of pumping blood throughout the body of creatures with spinal columns.

Furthermore, the physical property that partially identifies vertebrate-hearts as Vertebrate-hearts is intimately related to their functional property. That a vertebrate-heart is made of cardiac muscles partially defines it as a Vertebrate-heart, and allows it to perform the function of pumping blood throughout the body of a creature with a spinal column. Thus the function(s) of objective kinds may also be defining feature(s) of what they are, especially because these function(s) are intimately related to their non-arbitrary physical properties.^[8] Although there are such things as artificial vertebrate-hearts, these are not Vertebrate-hearts. They have the function of vertebrate-hearts but fail to have the non-arbitrary physical properties that partially constitute what vertebrate-hearts are. Thus they are not members of the objective kind Vertebrate-heart since they fail to be constituted by cardiac muscles.

2.2 Subjective Kinds

A subjective kind is defined here as a class of things that necessarily depends on human conceptualizations in order to constitute a unified category. Scarantino (2012) might refer to these as ‘ontologically dependent’ categories. Human conceptualizations typically involve the attribution of arbitrary properties, including functional properties, in order to identify things as members of a subjective kind. Thus a subjective kind may have members that are, using Barrett’s terminology, “ontologically objective” or “ontologically subjective” things (Barrett 2012, p. 417). What makes a subjective kind a subjective kind is some arbitrary, stipulated, conventional, or merely operational definition that serves to unify the members into a single class. For example, as mentioned previously, chairs are subjective kinds. What makes a chair a Chair is not its physical constituents, but rather that it has a certain structure (e.g. a seat and legs of some kind) and is conceived of as having the purpose of being sat on. This also does not imply that physical constituents are not necessary for members of subjective kinds to exist. A chair needs to be made of something in order for it to exist, but the physical properties of a chair do not contribute to the chair being identified as a Chair. A chair can be made of steel, plastic, straw, etc., but that it is made of any of these materials does not identify a chair as a Chair. An institution, such as the International Astronomical Union (IAU), is also a subjective kind. Although it is constituted by a collection of people, a mere aggregate of people does not constitute the IAU. All institutions depend on its members having a common purpose, which is necessarily determined by human conceptualizations, and it is around this purpose that the actions of its members are organized. One common purpose shared by members of the IAU is to specify what counts as a Planet.

Subjective kinds may also have physical properties that define them as the kind of thing they are. For example, a planet has physical properties that define it as a Planet, but it is an arbitrary fact – based on human conceptualizations – that these physical properties define what planets are. Thus the fact that Pluto is no longer a planet is not a discovery since it no longer being a Planet was established by fiat, which necessarily involves human conceptualizations. The property of being manufactured or man-made, however, is not necessarily indicative of subjective kinds. For example, diamonds and manufactured diamonds constitute the objective kind category Diamond. They are unified by the non-arbitrary, physical property of carbon atoms bonding into tetrahedral units. They also have functional properties that are intimately related to this physical property; nevertheless, these functional properties do not define them as what they are since the functional properties necessarily depend on human conceptualizations. For example, because of their molecular structure, diamonds have the property of being the hardest things known to exist. Thus they are used for making precision cuts, but this function is not what defines Diamond as an objective kind.

3.1 Optimists about Ordinary Language

Optimists understand folk emotion terms to be trans-theoretical terms in relation to the language of their theory (e.g. Averill 1980; Lyons 1980; Armon-Jones 1986; de Sousa 1987; Lazarus and Smith 1990; Ekman 1992; Solomon 1993; Levenson 1999; Wierzbicka 1999; Ben-Ze’ev 2000; Goldie 2000; Nussbaum 2001; Izzard 2009; Scherer 2009; Barrett et al. 2015; Ortony and Clore 2015). They understand themselves as referring to the same sorts of things as speakers of ordinary languages. Optimists hold that the concepts associated with folk emotion terms in ordinary languages are approximately correct definite descriptions, especially in the sense that they refer to interestingly unified sets of phenomena. This does not entail that according to optimists there is a one-to-one correspondence between a specific emotion term and a specific physiological state, neurophysiological state, or a unique set of these states. It also does not entail that there is a one-to-one correspondence between a specific emotion term, and a specific thought or behavioral pattern. Nor does it entail that emotion terms have a one-to-one relation with a category of objective kind or a category of subjective kind.

Optimists also accept, either implicitly or explicitly, that there is a reciprocal relation between ordinary languages and scientific languages. Theorizing begins with folk emotion terms, which reflect folk intuitions, and the process of theorizing aims at elucidating, clarifying, and correcting the folk concepts that are associated with folk emotion terms. These clarifications and revisions are not understood as independent theoretical achievements, but as achievements of a society as a whole. Thus optimists typically accept that folk emotion terms in Ordinary Language and the same terms in Scientific Language are associated with the same referents. They liken ordinary folk emotion terms to terms like ‘water,’ where earlier folk concepts may have been imprecise, broad, or even incorrect to a certain extent, yet they were still approximately correct definite descriptions or at least accurately referred. For example, when researchers working with the Aristotelian conception of water eventually discovered that water is H₂O, this reconceptualization did not entail a change in reference.^[12] Thus, according to the optimistic perspective, when the contemporary concept of water replaced the older Aristotelian concept in academic communities, as well as in the community at large, no linguistic shift would have occurred. The reason being that these terms were understood to be trans-theoretical terms.

3.2 Pessimists about Ordinary Language

Pessimists deny that folk emotion terms are trans-theoretical terms in relation to the language of their theory. According to pessimists, pessimists are not speaking of the same sorts of things as speakers of ordinary languages when pessimists use their emotion terms or redeploy folk emotion terms in the language of their theory. Pessimists typically argue that there is no significant relation between the meanings of folk emotion terms in ordinary languages and the same terms in their theory (e.g. Sarbin 1986; Griffiths 1997; Prinz 2004; Charland 2005; Panksepp 2008; Russell 2009; Scarantino and Griffiths 2011). They often ground this conclusion by arguing that most, if not all, of the concepts associated with ordinary language emotion terms are not approximately correct definite descriptions. They suggest that most, if not all, folk emotion concepts are more incorrect than correct.^[13] Pessimists also take the prescriptive stance that scientistic research – research that specifically involve the approaches of only the natural or social sciences – should not rely on the redeployment of folk emotion terms since these terms are typically thought to suffer from reference failure.^[14] Accordingly, they suggest a revisionist strategy for scientistic pursuits. They propose that those involved in scientistic pursuits abandon or radically alter the use of folk emotion terms in order to adopt a new scientific language. Given this prescribed disassociation with folk emotion terms and their associated meanings, pessimists hold that folk emotion terms and scientific emotion terms are not or ought not to be understood as trans-theoretical terms. They typically suggest that these two categories of emotion terms are or ought to be understood as polysemous terms.

4.1 Realism about Emotion

Realists are optimists about ordinary languages (e.g. Lyons 1980; de Sousa 1987; Lazarus and Smith 1990; Ekman 1992; Levenson 1999; Ben-Ze’ev 2000; Goldie 2000; Izzard 2009; Scherer 2009). They answer, “Yes” to the question of what Emotion is from the meta-semantic perspective. They claim that there is an important relationship between folk emotion terms and scientific emotion terms. More specifically, clarifying the meaning of folk emotion terms is thought to be a central task of emotion research. Scientific emotion terms and concepts are corrected, clarified, or further elucidated versions of folk emotion terms and concepts. Thus realists regard folk emotion concepts to entail some approximately correct definite description, and they regard folk emotion terms to be trans-theoretical terms. The justification for understanding the relationship between Ordinary Language and Scientific Language is that both types of languages share the same referents. From the metaphysical perspective, realists claim that emotions constitute an objective kind. They hold that all emotions are unified by some objective, non-arbitrary, physical property or set of properties (including the property of causal efficacy) that can possibly exist independent of human conceptualizations. These essential elements are typically characterized in terms of innate, biologically given emotional mechanism(s) or system(s), which are thought to be responsible for the cohesion of various elements of an emotional experience. Thus realists generally understand emotions to involve significant causal connections between the psychological (perceptions, thoughts, appraisals, etc.), neurophysiological, feeling, expressive, or behavioral aspects of emotional experiences.

4.2 Instrumentalism about Emotion

Instrumentalists are also optimists about ordinary languages (e.g. Averill 1980; Armon-Jones 1986; Solomon 1993; Nussbaum 2001; Colombetti 2009a,b; Barrett et al. 2015).^[17] They also answer, “Yes” to the question of what Emotion is, understood from the meta-semantic perspective. They hold that folk emotion terms are or ought to be understood as trans-theoretical terms. The meanings of folk emotion terms are typically taken to be approximately correct definite descriptions of what emotions are, even if such concepts need further correction, clarification, or refinement. Instrumentalists also ground their emotion terms in the same referents as ordinary language emotion terms. From the metaphysical perspective, instrumentalism denies that emotions constitute an objective kind. They deny that there is some objective, non-arbitrary, physical property, or set of properties that are possibly independent of human conceptualizations, that unify all emotions. What unifies all emotions into a single category, according to instrumentalists, is that they involve a unique class of conceptual categories, such as judgments, appraisals, social roles, social functions, or the like. These conceptual categories are necessarily dependent on human conceptualizations.

4.3 Eliminativism about Emotion

Eliminativism is not a common position among emotion researchers compared to realism and instrumentalism about Emotion (e.g. Sarbin 1986; Russell 2009).^[18] This is because theorizing about Emotion has typically been carried out under the opposing assumption of optimism, and eliminativists are pessimists about ordinary languages. Eliminativists answer, “No” to the question of what Emotion is from the meta-semantic perspective. They understand folk emotion terms to have meanings that are inadequate for the pursuits of scientific research. This understanding of the meanings of folk emotion terms is rooted in two interdependent assumptions: First, it is assumed that the meanings of folk emotion terms entail that Emotion or each species of Emotion (e.g. anger, joy, sorrow, etc.) is an objective kind. Second, it is concluded that folk emotion terms suffer from reference failure because there is no evidence to support positing Emotion or any species of Emotion as an objective kind. From a metaphysical perspective, eliminativists claim that Emotion is a subjective kind. Thus not only is it the case that folk emotion terms fail to refer, but it is not possible for them to refer since eliminativists take folk emotion terms to refer to objective kinds. According to eliminativists, there is no non-arbitrary, physical property that unifies all of our emotional experiences into a single coherent class. Furthermore, given that folk emotion terms have problematic meanings, and Emotion is not an objective kind, eliminativists also propose that the Language of Science be revised so as to eliminate folk emotion terms and concepts from the scientistic discourse on emotion. They do not, however, suggest the elimination of ordinary language emotion terms and concepts for non-scientistic endeavors since they consider the aims of scientistic research to be different from non-scientistic endeavors.

4.4 Eliminative-Realism about Emotion

Eliminative-realism is also an uncommon position compared to realism and instrumentalism (e.g. Griffiths 1997; Prinz 2004; Charland 2005; Panksepp 2008; Scarantino and Griffiths 2011, Scarantino 2012). This can be attributed to the fact that eliminative-realists, like eliminativists, hold that folk emotion terms are not or ought not to be understood as trans-theoretical terms. Thus eliminative-realists also answer “No” to the question of what Emotion is from the meta-semantic perspective. They are pessimistic about folk emotion terms and concepts. Like eliminativists, eliminative-realists differentiate scientistic research on Emotion from research on folk emotion terms and concepts. They hold that the study of folk emotion terms and concepts does not aim at obtaining truths about objective kinds, which is taken to be the proper aim of scientistic research. They also do not believe that folk concepts are approximately true definite descriptions of what emotions are. They argue that folk emotion terms and concepts entail that Emotion as a class, along with each species of Emotion (e.g. anger, joy, sorrow, etc.), is an objective kind, and they offer evidence to suggest that ‘emotion’ and other emotion terms refer to various disparate kinds. Therefore, eliminative-realists also argue that folk emotion terms suffer from reference failure, and propose that they ought to be eliminated from the languages and conceptual frameworks of scientistic research. From the metaphysical perspective, eliminative-realists, like realists, understand Emotion to be an objective kind. They hold that there is at least one non-arbitrary, physical property that is possibly independent of human conceptualizations that unify all or some emotions into an objective kind. Typically, this physical property is a property of a biologically given mechanism or system of Emotion. This mechanism or system might also be understood in terms of causal relations that are imposed by an Emotion mechanism or system, which explains the cohesion between the psychological (perceptions, thoughts, appraisals, etc.), neurophysiological, feeling, expressive, or behavioral aspects of emotional experiences.^[19]

5.1 The State of Crisis in the Science of Emotion

If it is true that the Science of Emotion is currently facing a crisis, we must be able to clearly identify the source of the problem in order to provide an appropriate solution. A brief survey of the various articles presented in the two special sections of Emotion Review that are dedicated to defining Emotion reveal the following consequences associated with the lack of a consensual definition of Emotion, the questionable nature of Emotion, and the meanings of folk emotion terms:

1. Fruitless debates

2. Major impediments to interdisciplinary dialogue and research collaboration

3. The misplaced tendency to separate cognition from emotion

4. Unfalsifiable theories

5. Blindness to cultural differences due to an emphasis on English words

One may argue that disagreements over the status of Emotion as a natural kind is what lies at the heart of these problems. For example, one may argue that an answer to the question of whether or not Emotion is a natural kind would yield testable hypotheses that could adjudicate between competing theories of Emotion. Such an argument, however, not only provides an analysis in terms of ‘natural kinds’ and ‘social constructions,’ which no longer clarify the discourse on Emotion, but also oversimplifies the problem by failing to explicitly distinguish the metaphysical perspective from the meta-semantic perspective. As suggested by the meta-semantic taxonomy of theories of Emotion, theories of Emotion can be distinguished by their metaphysical perspective as well as their meta-semantic perspective. Thus answers from these two perspectives can logically, conceptually, and metaphysically come apart.

5.2 Clarity Through the Meta-Semantic Taxonomy of Theories of Emotion

In regard to the natural kind status of Emotion, although both the metaphysical and meta-semantic perspectives are ultimately concerns regarding the nature of the referent of the term ‘emotion,’ the meta-semantic taxonomy helps us see that there are in fact at least four possible answers to the question of what Emotion is rather than the two answers that “Emotion is a natural kind” and that “Emotion is a social construction”:

1. Realism – The ordinary language term ‘emotion’ refers to an objective kind.

2. Instrumentalism – The ordinary language term ‘emotion’ refers to a subjective kind.

3. Eliminativism – The scientistic term ‘emotion^S’ refers to a subjective kind.^[20]

4. Eliminative-realism – The scientistic term ‘emotion^S’ refers to an objective kind.

These four possibilities suggest that the metaphysical and the meta-semantic perspectives are logically and conceptually distinct. Furthermore, even if we grant the possibility of establishing whether or not Emotion is an objective kind or a subjective kind, the attempt to do so fails to recognize that such an answer requires us to presuppose an answer to the question of what Emotion is from the meta-semantic perspective. It would require us to presuppose whether the appropriate domain for the kind of scientific research that we are concerned with (viz., Emotion) is the domain of ‘emotion’ or ‘emotion^S.’ Thus two out of the four competing categories of theories of Emotion would be privileged, albeit circularly, above the other two regardless of the status of Emotion as an objective kind or a subjective kind. Consequently, concerns from the metaphysical perspective are logically and conceptually distinct from the meta-semantic perspective, and vice versa.

Furthermore, concerns regarding what Emotion is from the meta-semantic perspective do not have any necessary metaphysical implications on concerns regarding what Emotion is from the metaphysical perspective. The reason why can be illustrated by the difficulties that arise in emotion research due to either the cultural diversity or cultural universality of Emotion concepts.^[21] In order for the cultural diversity of emotion concepts to have any metaphysical import on the status of Emotion as an objective kind, it must be shown that the cultural diversity of emotion concepts is significantly related to concerns regarding the objective kind status of Emotion. In other words, one must prove that one of the following is the case: 1) that the cultural diversity of emotion terms or concepts implies that Emotion is not an objective kind or 2) that the fact that Emotion is an objective kind implies that emotion terms or concepts are not culturally diverse. I respectively refer to these two statements as CD^E and OK^E:

• CD^E: If emotion terms or concepts were culturally diverse, then Emotion would not be an objective kind.

• OK^E: If Emotion were an objective kind, then emotion terms or concepts would not be culturally diverse.

These two possible relations, and their logical equivalents, exhaust the possible ways in which the cultural diversity of emotion terms or concepts can be logically and conceptually related to the metaphysical status of Emotion as an objective kind. Whether or not the cultural diversity of emotion terms has any metaphysical import on the metaphysical status of Emotion depends on the truth of at least one of these statements. Some pessimists have attempted to argue against optimism by assuming the truth of CD^E. I argue here that CD^E is false.

One might suggest that nothing we know suggests that either OK^E or CD^E is false. However, there are analogous cases of culturally diverse terms and concepts being associated with objective kinds. For example, the cultural concept(s) associated with the English folk term ‘cow’ and the cultural concept(s) associated with its cognate in Hindu Sanskrit ‘gow’ or ‘gau’ are very different, yet cows constitute an objective kind (Bovine). Reasoning by analogy, it would follow that CD^E would be false: it would be false that if emotion terms or concepts were culturally diverse, then Emotion would not be an objective kind. Furthermore, one might suggest that given this conclusion, it would logically follow that every rational emotion researcher believes in the truth of the proposition that would make statement OK^E false. This is because statement CD^E is logically equivalent to statement OK^E by the logical rule of inference referred to as transportation, which states that statements of the kind “If p, then q” are logically equivalent to statements “If ~q, then ~p.” In other words, statements that have these forms, given the same contents, will have the same truth-values. So if one statement were false, then the other statement would also be false by logical necessity, and vice versa. Thus, given the validity of the rule of transportation and the rational norm of logical consistency, it should follow that everyone who believes that CD^E is false ought to also believe that OK^E is false, and vice versa. Finally, those who hold some other belief about the truth-values of CD^E and OK^E would bare the burden of proof to show that their beliefs are either not false or not irrational in some appropriate way, which includes the condition of not being inconsistent or running into vicious circularities.

This argument addresses the lack of metaphysical dependence between the cultural diversity of emotion terms, or concepts, and the objective kind status of Emotion by illustrating how concerns regarding the cultural diversity of emotion terms or concepts have no metaphysical import on what Emotion is. We can draw the same conclusions, mutatis mutandis, about species of Emotion and any other postulated objective kind. Furthermore, we can draw a corollary conclusion about the relationship between conceptual facts and metaphysical facts: Conceptual facts do not entail any metaphysical facts, although the goal of “getting concepts right” might be to ensure that such entailments pertain. As to whether or not metaphysical facts entail any logical or conceptual facts, I leave open to further investigation and arguments. For example, although the cultural diversity of emotion terms or concepts does not have any metaphysical import on what Emotion is, the cultural universality of emotion concepts might (see Wierzbicka 1999; Prinz 2004). Thus the resolution to the crisis in the Science of Emotion may not lie in determining the natural kind status of Emotion nor in demarcating an aspect of the Science of Emotion, specifically the Folk Emotion Project, as being outside its area of concern.