Abstract
This article addresses Simon Lohse’s and Daniel Little’s responses to my article “Is Social Ontology Prior to Social Scientific Methodology?.” In that article, I present a pragmatic and deflationary view of the priority of social ontology to social science methodology where social ontology is valued for its ability to promote empirical success and not because it yields knowledge of what furnishes the social world. First, in response to Lohse, I argue that my view is compatible with a role for ontological theorizing in the social sciences. However, the view that results instrumentalizes social ontology. Second, in my response to Little, I argue that the same considerations I made in my article apply to naturalistic attempts to motivate a non-deflationary view, repeating some of the central issues of that article.
1. Introduction
In this article, I respond to Simon Lohse’s and Daniel Little’s responses to my article, “Is Social Ontology Prior to Social Scientific Methodology?” (Lauer 2019). Section 2 presents the main ideas behind Lauer (2019). Section 3 addresses Lohse’s response and argues that, contrary to Lohse’s position, what I call the “pragmatic view” is compatible with a role for ontological investigations in social science, but this role leaves its value as purely instrumental. Section 4 addresses Little’s response and his naturalistic social ontology. I argue that we can rehearse the arguments of Lauer (2019) for the positions he carves out and adopt a more theoretically modest view than his naturalistic social ontology. Section 5 concludes.
2. Lauer (2019)
In Lauer (2019), I address what I call the ontology matters! (OM!) thesis:
Ontology contributes to the success of science (success in explanation and/or prediction) because answering ontological question can contribute to empirical success, i.e. successful predictions and/or explanations.
I present the possibility, following Hofweber (2009, 2016), that there are multiple ways in which ontological questions might be answered (that fall in line with different ways of thinking about existential quantification). If social ontology aims at the answering of social-ontological questions, then we may formulate at least two views that result in different kinds of OM! arguments:
Pragmatic View (PV): Social ontology can aid in empirical success by introducing statements into our social scientific theories/models that make them empirically adequate. These statements do not entail a commitment to what there is. (Lauer 2019, 183)
Realist View (RV): Social ontology can aid in empirical success through the successful answering of ontological questions. Social scientific theories and models become empirically adequate by determining what there is (using the methods needed for doing social ontology; Lauer 2019).
As I define it in the article, by “social ontology” I mean analytic social ontology, an area engaged in conventional kinds of philosophical inquiry. This usually involves the careful analysis of concepts, the use of thought experiments, and the pumping of intuitions about a variety of social entities, for example, the large literature on group and collective agency. 1 OM! arguments give a place to social ontology as an area of pure philosophical investigation in the methodology of science. RV allows the traditional conception of this discipline to have that place.
PV is the view we arrive at if we assume that answering ontological questions involves the use of quantifiers that do not imply ontological commitments (their truth, instead, is internal to the linguistic frameworks we deploy). If answering ontological questions does not entail ontological commitment, then social ontology would seem to contribute to empirical success by affording us with theories/models that allow scientists to achieve their pragmatic and/or epistemic goals. The general idea is that social ontology, by answering questions about what falls within the domain of our linguistic frameworks, may enable inferences that can help to promote empirical success in social science.
As Lohse rightly notes, I favor PV. In fact, I argue that where it comes to understanding the reasoning of social scientists: (a) PV and RV seem to be empirically equivalent, and (b) PV requires fewer philosophical commitments. Thus, the discussion to come will focus almost exclusively on PV.
In what follows, I shall present and examine Lohse’s and Little’s responses to my article. In responding to their challenges and positions, I develop further the position I set out in Lauer (2019). I carve out space for two positions:
Social ontology can be understood as a kind of modeling practice, a production line for “use and abuse” (French and McKenzie 2012) by social scientists. On this account, the value of ontological investigations can be understood in terms of their role in promoting empirical successes. The elements of these frameworks may be true in minimalist or deflationary senses.
The prospects for naturalized social ontology are unclear, but because we can understand the truth of ontological claims in minimalist/deflationary terms, we can take seriously social scientists’ ontological claims about the social world without requiring ontological commitment to the posits of social scientific theory.
I present both theses as a result of my attempt to reconcile PV with the objections of my interlocutors.
3. Lohse on the Value of Ontology
This section responds to Lohse’s challenge to my article and offers a critique of the alternative he formulates to PV. Lohse contends that the PV I advance denies the relevance of ontological investigations tout court (he treats it as an instance of an anti-ontological pragmatism).
Contrary to Lohse, I argue that PV is compatible with a certain way of thinking about ontological investigations. Ontological investigations are a natural part of what, taking inspiration from French and McKenzie (2012), we may call the production line view of social ontology. According to the production line view, social ontology contributes to empirical success in science not by shedding light on the nature of the world, but by supplying frameworks that can aid achieving empirical success. In other words, if we should adopt PV, then OM! arguments suggest that ontology is better imagined as a production line for work in the various sciences (and social ontology for social science).
3.1. Lohse’s Challenge
Lohse addresses my pragmatic position as follows: According to Lauer, the only task left [to social ontology] is to provide the social sciences with (potentially fruitful) statements that enable new explanations and predictions. But what would that really mean other than thinking of new assumptions to be tested in light of empirical evidence? It seems that this would not entail any kind of ontological reasoning (it is just what social scientists do in their day to day research). (Lohse, 2020)
In the passage above, Lohse takes the position, at least as I articulated it in my article, to imply that we should do away with ontological investigations tout court. This would make it a kind of anti-ontological pragmatism (Lohse 2017). Lohse’s argument seems to proceed by noting that if PV is true, then ontological investigations are not needed for successful science. However, it does seem that ontological investigations are important for the practice of science. Therefore, PV is false, social ontology can make contributions in a different way than what PV implies.
Contrary to what PV implies, we can conceive of the contributions of ontological reasoning to projects in the philosophy of the social sciences as in line with the philosophy of science in practice, which Lohse characterizes as follows: . . . philosophers of science in practice analyse scientific theories alongside epistemic practices, such as experimentation in the laboratory, explanatory modelling, and classification practices. This re-orientation introduces new kinds of questions to philosophy of science, which in turn suggest more naturalistic ways of answering them: How do the material aspects of the laboratory interact with experimental design? To what extent does big data research influence scientific classifications in biology? What are the implicit ontological assumptions of explanatory network models in sociology? (#)
I shall argue that the ontological questions addressed in Lohse’s argument are, in fact, compatible with PV. Instead of thinking about these questions in their full-blooded ontological sense, we might instead imagine them as questions about the linguistic frameworks chosen by scientists for their empirical tasks. For example, we might ask questions about the role of big data in classification without positing that the existence claims made in specific systems of classification refer to species of organism. Moreover, we might ask about explanatory network models in sociology and restrict our inquiry to existential claims within these models without reference to whether these models really do refer to elements of their target systems.
The success of Lohse’s response depends on what precisely it is that ontological reasoning (or, to be consistent with Lohse’s language: an ontological investigation) amounts to. As Lohse puts it, the PV is thoroughly anti-ontological. It implies that no ontological reasoning should bear at all on the practice of social science.
So, what is an ontological investigation? Lohse defines it as follows: The phrase “ontological reasoning” (or “ontological investigation”) should not be understood in the restricted sense of “the study of things as they really are” here, of course, as this would be tantamount to the realists’ approach. My use of the phrase is more innocuous. It is meant to refer to all kinds of analyses of ontological aspects and assumptions of the sciences. (#)
There are two definitions the above passage. The first definition Lohse mentions, according to which ontological investigations/reasoning lead(s) to conclusions about what really exists, is incompatible with PV. Because PV takes it that existential quantifiers used by social ontologists needn’t be interpreted as carrying an ontological commitment, it would follow that this form of ontological reasoning is beyond the scope of the job given to social ontology by PV.
However, the second definition, the definition endorsed by Lohse, is less obviously in conflict with PV. This defines an ontological investigation as any analysis of an ontological aspect or assumption of the sciences. Here is an unpacked version of Lohse’s definition, by my reading:
Ontological investigations are investigations of the inferential properties of a linguistic framework, 2 particularly those that can be understood as existentially quantified formulas or that concern individuals assumed by that framework.
The sorts of investigations that Lohse describes might then be taken to be relevant to understanding the “properties” of our linguistic frameworks. To understand those frameworks in part means looking at their social contexts in the practice of scientists. Unpacking the elements of these frameworks may involve picking out those assumptions that are only implicit in scientists’ day-to-day work.
But note that PV places emphasis on empirical success at the cost of ontological commitment. Presumably, our frameworks have certain properties (they entail certain consequences that we can arrive at by applying rules of inference), but they do not entail that the elements of the domain exist. Assuming that certain systems have certain properties may enable insights into those systems that are empirically fruitful. If we assume that a system has certain properties, then we may be able to infer that it has some other (perhaps observable) properties as well. Similarly, analytic social ontology as exemplified in the work of Epstein (2015), Searle (2008), and numerous others may provide different ways of representing social systems that are of interest to social scientists. In turn, these ontologies may allow for novel explanatory or predictive inferences. However, investigating systems under the assumption that they have certain properties (e.g., a group of individuals as if it is an agent) does not require that we commit to the existence of systems that are the way we assume them to be. In other words, whatever frameworks are offered to us by social ontologists, by PV, we do not need to understand them as really getting at what the social world is like. As long as ontological investigations are defined as above, and “ontological assumption” is not understood in its strict realist sense, then PV and OI are logically compatible.
What might the general picture I am describing look like in the actual practice of science? While I don’t know of good cases where the fruits of analytic social ontology have been put to use for attaining empirical success, some theoretical work in economics offers some insight into what the general strategy might be. We see this mode of reasoning vividly in models of stock market crashes developed by econophysicists (Sornette 2003). To represent stock market crashes, some econophysicists argue stock markets behave as critical systems—or, systems that undergo phase transitions. This involves the use of models from physics and so requires us to consider some general questions about the ontology of stock markets (particularly compared to the picture of the stock market presented in the Efficient Market Hypothesis). In turn, the behavior of these systems can be characterized by certain curves depicting distinctive rates of growth that occur in critical systems as they reach a critical point and undergo a phase transition. In this case, treating a system as a critical system allows scientists to connect systems investigated to empirical regularities (rates of growth) that a wider class of systems obeys. By adopting the framework that treats a stock market as a critical system, new insights are possible that might be precluded by other frameworks.
But, of course, there are deep theoretical questions to ask about whether, in fact, these models should be interpreted literally. This is made especially prominent by the fact that there are ways in which critical systems behave not exhibited in stock market systems, for example, different stock markets exhibit different critical exponents, but critical systems investigated by physicists are often characterized by the same critical exponents (Jhun, Palacios, and Weatherall 2017). 3 So, while these models might be useful for predicting or accommodating data about the stock market, there are significant questions to ask about the fit between the models and their target (in the event that we give these models a realistic interpretation).
In this case, we can see that there are ontological questions we can ask and pursue answers to: Is it true that stock market systems are critical systems? We can answer this question by examining the deductive consequences of the equations that describe those systems and the fit of the data relative to the predictions of those equations. We can consider both the data the equations originally intended to apply to and stock market data to determine whether, in fact, the use of these models is fruitful and the extent of the analogy between physical and economic systems. In this case, the slippage between the behavior of critical systems and stock market systems suggests that some careful hedging is important. If the value of these models is seen not in terms of their ability to describe stock market systems as they really are but instead purely in terms of the inferences they allow (the explanation and prediction of the data along oscillating logarithmic curves), then we can readily account for their success without raising worries about the strict literal interpretation of the models. 4 Whether stock markets are critical systems properly speaking is something to which we may demur.
My suggestion in Lauer (2019) is that we should be similarly cautious in our interpretation of the value of social-ontological frameworks for promoting empirical success. Where social-ontological frameworks are introduced, their contribution to empirical success does not need to be interpreted any further than in terms of what they contribute to empirical success or the practice of science more broadly construed. It is in this sense that we can say that the elements of these frameworks are “true.” However, there is, possibly, a benefit to thinking about such matters. Constructing frameworks for formulating hypotheses that might inspire testable social scientific hypotheses certainly seems like a useful task and it seems hasty to conclude that ontological frameworks are never useful in this way. However, the proof for a given ontological framework, by PV, is in the pudding.
In sum, OI and PV are compatible. PV does not imply that we should do away with investigations as characterized by OI. PV only states that answers to ontological questions need not involve any appeal to what there really is, even though these answers may have the form of existentially quantified formulas. If so, then we should think of the contribution of social ontology to empirical success in social science not in terms of its ability to inform us of what furnishes the social world, but instead in terms of how it contributes to social scientific goals. I allow that social-ontological theorizing might contribute to empirical success, but what I question is how it does.
4. Little’s Naturalism
Little’s response to my article is focused less on the logical implications of PV and more on the assumed definition of social ontology. Recall: my focus is on analytic social ontology in Lauer (2019). This is a field comprised of numerous philosophers developing complex ontologies directed at answering questions about, for example, whether there are group agents, whether there are institutions, the grounding conditions for facts about the social world, among other things.
Little is skeptical that pure philosophical questions about the ontology of the social world are fruitful. Little writes, I do not believe there is a field of pure philosophical reasoning within which philosophers can answer the question, “What exists in the social realm?” and where the answers have the status of substantive non-empirical truths.” (Little, 2020)
With this, Little expresses his disagreement with a conception of social ontology as a pure philosophical project directed at questions about the contents of the social world. In particular, the traditional conception of social ontology that Little has in mind is largely focused on a priori methods of answering these questions, which we can construe as the use of conceptual analyses, intuitions, and thought experiments to establish truths about what there is. Little does not see this as plausible and so sees a realist take on the results of social ontology as implausible, inasmuch as the methods in question are a priori. As such, Little rejects realist OM! arguments.
From here, Little presents two positions. They are as follows:
The inadequacy of Realist OM! arguments does not entail that ontological theories are not-veridical/truth-bearing.
Social ontology can be conceived of as an a posteriori discipline, one where ontological claims are offered as defeasible claims about the nature of the social world (grounded in empirical considerations).
The first of these occurs as a way of responding to how I present the conceptual terrain as consisting of realist and pragmatic OM! arguments. If realist OM! arguments fail, then the only alternatives (that I provide) are pragmatic OM! arguments (as characterized by PV). But, Little suggests, PV seems to suggest that ontological theories are not truth-bearing. The second, the positive proposal, is motivated by explicit appeals to Quinean approaches to generating ontological commitments. In this, Little seems to adopt a kind of naturalism about social ontology.
In what follows I respond to both claims. First, I argue that PV is not committed to denying that ontological claims are true. If my argument is sound, then this helps to preserve the adequacy of my distinction. Second, I argue that the naturalistic strategy faces significant and important challenges in the context of the social sciences, especially considering a broadly Quinean methodology for establishing ontological claims. I argue that the broadly Carnapian approach to thinking about social ontology exemplified in PV is preferable given these challenges.
4.1. PV and the Truth of Ontological Theories
PV presents a position that, as I intend it to be interpreted, is more flexible than Little’s claims suggest. The idea behind PV is only that ontological claims (claims that have the form of an existentially quantified formula) need not imply that the elements of their domains exist. However, we should not infer that because the existential quantifier need not entail an ontological commitment, that we should deny the truth of existential claims.
To show how we can make this idea intelligible, let’s consider some of the realist/anti-realist intuitions underlying this idea. Fine (1984) expresses these positions well when he characterizes what he calls the Natural Ontological Attitude—the natural ontological attitude is the common core that scientific realists and anti-realists share—both might acknowledge the electron as a posit of physical theory but at the same time have differing attitudes about the whether we really should believe that there are such things. The realist may stamp her foot and insist upon the reality of the electron while, similarly, the anti-realist would deny that we really need to believe such things. On Fine’s approach, we can draw a distinction between the claim that a theory is true, in the sense that we should believe in the literal truth of ontological claims it posits, and whether the thing exists.
PV allows for a minimalist conception of truth like Fine’s. To say that there are Xs when it is true is to say that it is true in an ordinary way. Consider Hofweber (2016)’s example: we may infer from the claim that Cheney admires Iago that there is someone that Cheney admires. Though the inference seems to rely on a premise that appeals to some entity we would withhold saying really exists (a fictional entity), this does not seem to render the argument unsound. Moreover, in response to the question of whether there is someone that Cheney admires, it would be correct to answer by noting that Iago is someone that Cheney admires. If we fix our domain to social scientific theories, we might urge the same. To say that there are races, social classes, markets, social structures, and so on, is just to assert truths in this ordinary way. That is, those parts of our social science that are accepted by their communities as bearing empirical significance are claims we can say are approximately true.
So, just as Fine advocates this minimalism as a way to take seriously scientific claims while eschewing realist commitments to the posits of scientific theories, I suggest the same for the social sciences.
This opens up the following position: the naturalistic standards according to which we say that social scientific theories are true do not entail that the theoretical entities of the social sciences exist (in the sense of RV). If a particular entity is indispensable for the prediction or explanation of some phenomena, we may regard claims including it as true in a deflationary sense. Though I don’t have the space to develop this position more fully (I aim to address this more completely in other work), this suggests that PV does not necessarily entail that ontological theories are not truth bearing. Rather, it restricts truth talk to a minimalist account of truth that relativizes truth talk to the frameworks and the acceptance of those frameworks by the relevant scientific communities. 5 PV, as I stated it in Lauer (2019), emphasizes prediction and explanation and deemphasizes realist commitments to highlight that it is the empirical role of social-ontological theory that vindicates it as useful or not for scientific theories. PV by itself does not require a denial that these theories could be true. However, to say that they could be true may not require that we believe that their posits exist in the sense relevant to realists and traditional social ontologists.
This view has implications for how we should think about naturalistic projects in social ontology. This brings us to the second of Little’s positions, let’s turn to consider this next.
4.2. Challenges to Naturalism about Social Ontology: OM vs. NOM?
Little articulates his position as follows: I argue for a third position on ontology that is different from either of the approaches described by Lauer—ontology without apriori reasoning, ontology as simply the outer limit of scientific and empirical theorizing, and ontology construed realistically as making claims that are approximately true . . . . I reject the possibility or utility of pure apriori reasoning in ontology. (#)
Given what I have argued so far, it seems that I agree with Little. In my responses to Lohse, I suggest that ontological investigations may be an important part of scientific practice (at least in the relevant sense defined above) and that PV does not reject that possibility tout court. Moreover, I present the possibility in the previous subsection that we can talk about truth without necessarily adopting ontological commitments. So, we would seem to agree because, per my earlier discussion, I claim that the truth of an ontological theory would depend upon the empirical evidence (or the empirical successes of the theory or model in question).
However, using these positions, we can draw further distinctions that reflect where Little and I disagree. We can separate normal ontology matters (OM!) arguments from naturalistic ontology matters (NOM!) arguments. I set them side by side so the reader can consider the differences.
Ontology contributes to the success of science (success in explanation and/or prediction) because answering ontological question can contribute to empirical success, i.e. successful predictions and/or explanations.
Naturalistic ontologies (ontologies grounded in the work of our best social science) contribute to the success of science because giving naturalistic answers to ontological questions can contribute to empirical success, i.e. successful predictions and/or explanations.
In turn, we might rehearse the arguments of Lauer (2019). The ontological claims of naturalistic ontologies may not require interpretation in realist terms. That is, the quantifiers used in the claims of our naturalistic social ontologies could be interpreted so that they do not require ontological commitments. The result is that we might distinguish between pragmatic and realist varieties of NOM! arguments. What could lead us to prefer pragmatic or realist versions of NOM! arguments?
I opt for conservatism and favor the pragmatic NOM! arguments. To see why, we should consider what would license realist commitments to ontological claims of our best social science. There is a burgeoning controversy about just this issue. Hawley (2018), for example, argues that the social sciences lack the relevant degree of empirical success to be able to use the no miracles argument, a standard tool of the scientific realist (Boyd 1980; Putnam 1975), to argue that we can believe that social scientific theories approximately describe what furnishes the social world. While there are certainly optimistic takes on naturalized social ontology (Kincaid 2016, 2018; Saunders 2020), the data so far suggest that the predictive track record of the social sciences leave the prospects for a realist and naturalist social ontology looking grim.
Nevertheless, Little may respond that prediction is not a good criterion for realism about the social sciences. In fact, Little argues for this position elsewhere (Little 2016). In that work, Little agrees that the social sciences are not as predictively successful as their natural scientific cousins. Nevertheless, Little disagrees that we ought to reject realism about the social sciences. He writes, If we believe that class conflict is a key factor in explaining political outcomes, we can do further sociological research to further articulate what we mean by class and class conflict, and we can investigate specific social and political processes to piece together the presence or absence of these kinds of factors. These investigations may give us confidence that “class” exists. (224)
Little’s view seems to be that explanatory factors can play a role in motivating realism about the social sciences. In this, he also seems to rehearse a view like Kincaid’s (2016, 2018), where he relies on appeals to the validity of constructs to motivate ontological commitments. On this view, numerous independent pieces of evidence can come together to validate a construct and thereby support claims to realism about the construct. Similarly, Little argues that we can bring together explanatory considerations that will raise our confidence in the existence of a social entity.
Does this argument succeed? This is unclear. Explanatory considerations seem to be more permissive than prediction. For example, explanatory considerations might have vindicated the existence of ether or caloric. In their time, they seemed to play the relevant explanatory roles in their theories and seemed to be vindicated by those roles. Yet, we would hesitate to say that explanatory considerations vindicate the existence of those entities. Accordingly, it seems that, following criteria for adequate ontologies like those traced out in Bryant (2017), explanatory considerations are not sufficiently robust. They allow too much theoretical content into an ontology. Predictive success would seem to be more robust and thus more restrictive of what theoretical content passes into an ontology. Moreover, we might align the disposition toward naturalistic ontologies with and aversion to epistemically risky existential claims (Chakravartty 2017). Claims with lower degrees of epistemic risk are statements that we are more confident that we can test and prove false. Prediction might fare better (for the naturalistic social ontologist) because theories that yield novel predictive success deploy lower risk statements and so serve as a superior basis for ontological commitment (commitments based on explanatory considerations would be more risky). These ideas suggest that, explanatory considerations might drive us away from naturalistic social ontology, not toward it.
This might suggest that I favor anti-realism about the social sciences. 6 However, I am moved by the Finean line given earlier. Rather than move to anti-realism, we may instead deny that we should be realists but not accept anti-realism. It seems important to vindicate the truth of claims made by scientists. By the same token, it is important to be skeptical where the evidence might warrant it. In light of this, I argue that we should favor pragmatic versions of NOM! arguments. 7
5. Conclusion
This article provided responses to objections presented by both Simon Lohse and Daniel Little to my article “Is Social Ontology Prior to Social Scientific Methodology?” The positions presented by Lohse and Little are valuable, help to trace out the conceptual space around this area that Lohse (I think aptly) refers to as “meta-social ontology,” and forced me to more carefully consider the details of the position I intended to advance in my original article. I hope that the views I articulated through my responses help to further enrich the space of positions that one might hold in this area. Through my responses to their objections, I formulated two positions:
Social ontology can be understood as a kind of modeling practice, a production line for “use and abuse” (French and McKenzie 2012) by social scientists. On this account, the value of ontological investigations can be understood in terms of their role in generating empirical successes. Elements of these frameworks, to the extent that we can say they are true, may only be true in minimalist terms (but their truth depends on whether they generate empirical successes).
The prospects for naturalized social ontology are unclear, but because we can understand the truth of ontological claims in minimalist/deflationary terms, we can take seriously social scientists’ ontological claims about the social world without requiring ontological commitment to the posits of social scientific theory.
So, the views formed in response to Lohse’s and Little’s critiques are twofold: an instrumentalism about the aims of social ontology, and a naturalist, though deflationary, approach to doing social ontology.
Footnotes
Acknowledgements
I would like to thank Simon Lohse for proposing the idea of putting together this discussion forum, Daniel Little and Harold Kincaid for agreeing to participate in it, and this journal for accepting the proposal. Simon and Daniel put together interesting and forceful points for me to consider and I was ecstatic to be afforded the opportunity to develop the position of my paper through this forum.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
1
For examples and overviews of this literature, see List and Pettit (2011) and 
.
2
Where I allow that sets of models or theories count as such frameworks.
3
Critical exponents play an important role in models of critical phenomena. These exponents drive the behavior of the system when it reaches a critical point—typically, as a system approaches a critical point, the system exhibits divergence (values generated from the equation begin to diverge). These critical exponents are stable above and below the critical point for a wide range of systems. However, in the case of stock market models, it seems that these critical exponents do not remain the same for different stock market systems, suggesting that stock markets do not comprise what physicists call a “universality class.”
4
One might be curious about my use of “explanation”—I intend it in a pragmatic sense consistent with the work of Achinstein (2010) and . These accounts allow that explanation, constrained by background theory and the explanation’s audience, might deviate from a strictly true account of a phenomenon.
5
Some notes here: this reasoning would cut against other arguments in extant work. For example, argues that a neo-positivist metaphysics would allow for commitments to the indispensable posits of scientific theories that have metaphysical significance for the community of scientists. My suggestion is that indispensability tells us something about the framework in question, less about what there really is.
6
Though, it should be noted: like many others, I am sympathetic towards selective realism and so think questions about realism should be answered with due deference to the relevant empirical evidence that can license ontological claims.
7
The resulting view bears resemblance to Neo-Carnapian ontology. While Thomasson refers to her view as “easy ontology,” we might call the view I present a version of “easy social ontology.” On this view, our social ontological commitments are settled by looking at the application conditions for terms used in our social scientific discourse and then to allow that “exists” may attach to terms about objects whose application conditions are satisfied.