Speak no ill of the dead: the dead as a social group

In her recent article “The Ontology of Social Groups”, Thomasson (Synthese 196:4829–4845, 2019) argues that social groups can be characterized in terms of the norms that surround them. We show that according to Thomasson’s normativity-based criterion, the dead constitute a social group, since there are widespread and well-defined social norms as to how to treat the dead, such as the norm expressed in the title (“Speak no ill of the dead”). We argue that the example of the dead must not be interpreted as a counterexample to this criterion, and that, rather, there are good reasons to think of the dead as a genuine social group. Furthermore, the dead as a social group exist, regardless of whether or not the dead (i.e. the members of this social group) exist as persons. This view is clearly incompatible with the idea that social groups can be explained by citing features of their members, which is why we take the case of the dead to pose a challenge for the thesis of individualism. We consider the ontological implications of this example and suggest that social groups ought to be seen as reifications of social norms. Another implication of taking the dead to be a social group is that the presence of external norms is sufficient for the existence of a social group, and therefore internal norms are not necessary. We defend this implication against some potential theoretical and moral problems.


The normativity of social groups
Amie Thomasson argues that it is possible to distinguish social groups from "arbitrary assemblages" by looking at the norms that govern these groups: There are no governing norms for how Genghis Khan and I are both to be treated, nor (in our culture) for how people with a longer second toe, or hazel eyes, are to be treated, or are to act.There are, unfortunately, such norms for how people of color, and women, are to act and be treated.As participants in the social world, we can be blind to these norms and fail to recognize the social groups […] only at our own peril -peril of sanctions, ostracism, or worse.(Thomasson, 2019, p. 4840) In order to navigate our social world, it is important to have the capacity to identify social groups only insofar that it is the case that identifying these groups allows us to recognize how members of the groups might act or might be treated.Social groups have social significance only insofar as they have implications as to how its members are to be treated or are to act.This is why it might be possible to characterize social groups by looking at their normative implications.According to Thomasson, there are three kinds of norms that characterize social groups: internal norms, external norms, and structuring norms.(Thomasson, 2019, p. 4838) Internal norms determine how the members of the social group are to act.For instance, members of a baseball team, qua being members of a baseball team, are expected to attend practice.By joining a baseball team, one subjects oneself to the norm to attend practice, which means that, regardless of whether one feels like attending practice, one has a reason to show up.External norms determine how members of a social group are to be treated.A familiar norm to many (perhaps all) cultures is that we are to treat the elderly with respect.Finally, structuring norms are those norms that give structure to the social group.These are the norms that, for example, give first-year students such a hard time in fraternities and sororities (while also making them turn on the next generations of first-years in the years after). 1  Thomasson argues that social groups do not always display all three kinds of norms.She also does not take any one of these norms to be more central or essential to the 1 We focus our discussion on conventional norms only.Articulated in terms of judgments, Nicholas Southwood classifies conventional norms in the following way: "Conventional normative judgements are essentially practice-dependent in the sense that they are necessarily grounded, at least in part, in presumed social practices" (2011, p. 781).While more robust normativity (reason-giving or otherwise) may be found in discussions of moral or rational norms, our treatment of norms is essentially "lightweight"-i.e., their authority is contingently grounded by current social practices.Without straying too far into metaethical territory, we do not see our notion of norms to be problematic.For, we are talking about norms akin to those of chess, law, grammar, and etiquette.And as John Brunero aptly points out, the "question of whether we always have reason to comply with the requirements of chess, law, grammar, and etiquette is up for debate.But the relevant point here is that we don't think that if that debate yielded a negative answer, we'd have to give up on the idea that these requirements exist" (2020, p. 210).Finally, we believe that our lightweight treatment of norms is appropriate to the discussion of social groups.For, this framework allows for coherent social debate over which norms are "right", "correct", "justified", etc.Perhaps the norms that we should hold or adopt will turn out to be robust in the sense of moral, rational, or epistemic norms.However, in order to discuss potentially unjustified norms (e.g.sexist, racist, or other problematic norms), this lightweight treatment of norms is useful.existence of social groups than the others.While she does believe that looking into the normativity of social groups grants us a criterion that helps us distinguish them from arbitrary assemblages, she is not trying to give an account of the nature of social groups, and thus is not looking for necessary and sufficient conditions.Thomasson resists giving such an account, because she believes that the term 'social group' is in a sense artificial and the chances of finding a unified encompassing account of social groups are slim.Thus, she proposes that instead of asking "What are social groups?"we must ask ourselves "What do we want the concept of a social group for?"(Thomasson, 2019, p. 4836).
This shift between the ontological question and the functional one is motivated by Thomasson's underlying conception of metaphysics.We would ordinarily hope that answering the question "What are social groups?" could shed light on the ontological status of social groups, and more specifically on the relation between social groups and the individual members that constitute them.However, Thomasson boldly dismisses the core ontological question whether social groups exist, on account of its answer being already obvious: of course social groups exist.
Thomasson is an advocate of what she calls the "easy approach to ontology" (Thomasson, 2015, p. 20).Her approach rejects the Neo-Quinean hard interpretation of metaphysics, according to which existence-questions are robust and we cannot answer them straightforwardly.On the contrary, Thomasson argues that existence-questions are easy, if not trivial, and that their answers are not based on metaphysical theorizing, but on conceptual analysis, competence in language, and empirical evidence.She believes that her easy approach to existence-questions allows for a reorientation of metaphysics towards new and more useful directions.Therefore, according to Thomasson, asking whether social groups exist would be rather silly, and we should shift the focus from the ontology to the function of social groups.
Given this metaphysical framework, Thomasson believes that those who deny that social groups exist cannot refer to them and thus cannot speak of them in ordinary discourse.(Thomasson, 2019, p. 4833) Since giving up on this vocabulary would make it nearly impossible to navigate the social world, rejecting the existence of social groups is a fool's game.In spite of possible expressivist or fictionalist attempts to recover social group talk, we agree with Thomasson that it would be foolish to believe that social groups, such as baseball teams, universities, discussion groups, etc., do not exist at all.However, we reject Thomasson's easy approach to ontology.We believe that there is an important and interesting sentiment expressed by those who argue that social groups do not "really" exist, which is that there are no such things as social groups over and above the individuals that constitute them.This sentiment is worth taking seriously, and Thomasson's argument, insofar as it is meant to stop this kind of thinking in its tracks, misses the mark.
We will not dwell too much on the reasons why Thomasson's deflationary approach is problematic, since it would lead us too far from the scope of the paper. 2However, though we disagree with Thomasson's metaontology-her rejection of existencequestions-we agree with her epistemological thesis-that it is easy to show that entities, such as social groups, exist via conceptual analysis or empirical evidence.Thus, social groups clearly exist, but there are nevertheless interesting questions to be asked about their metaphysical status.To this effect, we follow Daniel Korman (2019) in making this distinction between the two theses and in only accepting the epistemological thesis.Korman argues that Thomasson's metaontological position rests on a contentious theory about the possible meanings of the word 'object', for it is useful to show that the two can be separated. 3Following this suggestion, we claim that Thomasson's ontology is not as "easy" as she would like it to be, and that some deeper metaphysical work is required.Indeed, we think that an ontological interpretation of Thomasson's normativity-based criterion has some very interesting metaphysical implications that one could only appreciate if one rejects Thomasson's metaontology.

Speak no ill of the dead
As we mentioned above, even though Thomasson's account of social groups only aims to give a characterization of social groups, not a story of their ontological status, we believe the reasons for resisting such a story are ill-founded.Furthermore, Thomasson's characterization can be used to give examples of social groups that pose an interesting challenge for individualist accounts of social groups. 4Amongst those is the leading example of this article: The Dead.The title of this paper is derived from the Latin phrase "De mortuis nil nisi bonum" which literally means "Nothing but good of the dead".The phrase conveys the social norm that we are not to speak ill of the dead because they are unable to defend themselves.The phrase is sometimes used more metaphorically about those who are absent, but more often than not, it is a norm that specifically applies to the dead.It is a clear-cut example of an external norm.Note, for instance, that it does not apply to people who happen to have died, but it applies to the dead qua being dead; it is not because Lucy was such a great person that we should not speak ill of her, but it is because she is dead that we must refrain from speaking ill of her.This indicates that those who have passed gain a peculiar social status, that is bestowed upon them solely because they are among the dead.We commonly take the dead to be deserving of our respect, regardless of who they were or what they achieved when they were alive.Of course, if they did truly terrible things this would counteract the general norms such as the norm to speak no ill.However, this is no different for members of any other social group.For instance, while we are generally to respect the 3 Thomasson claims that there are only two possible meanings of the word "object", a sortal meaning and a covering meaning.(Thomasson, 2015) Korman argues that there is a further possible meaning of "object", which is an unrestricted one.According to Thomasson, the sortal and covering meanings are the only ones that we need in metaphysics, and they both lead to eliminativism with respect to existence-questions.On the other hand, Korman shows two key points: (1) the unrestricted meaning is a necessary tool to clarify debates about the other two meanings, and (2) the unrestricted meaning restores the importance of existence-questions. (Korman, 2019). 4Individualism about social groups is the thesis that every aspect of a social group can be explained citing only the features of the individual members of the social group.We will explain this more extensively later on.
elderly, we are not obliged to respect every single individual belonging to this group, and nor do we. 5 Since we have various norms as to how to treat the dead, Thomasson's normative characterization of social groups tells us that the dead constitute a social group.Opponents of this characterization might want to try to cite this example of the dead in order to show that the criterion of normativity, on which the characterization is built, is defective, since The Dead "clearly" cannot be a genuine social group.However, this is not at all clear to us.The underlying argument seems to be that the dead do not exist-not as persons at least-and something could only be a social group if it has members that are existing persons.
There are many possible views concerning the metaphysics of the dead.It might seem at first that some of these views are clearly incompatible with the idea that The Dead is a social group.We argue to the contrary, that all commonly endorsed views on the metaphysics of the dead are in fact compatible with the existence of The Dead as a social group.From here, we can formulate an important lesson, which is that the existence of social groups should be taken to be distinct from the ontological status of the members of these social groups.
There are roughly six different kinds of views on the metaphysics of the members of The Dead that are common enough to take into consideration. 6These are the views: 1.Members of The Dead are active persons.If (1) were the case, then The Dead would uncontroversially be a social group.However, (1) would presumably only be believed on religious or spiritual grounds.The second view is that members of The Dead are inactive persons.This view aligns with Patrick Stokes' position.He cites the various social norms surrounding The Dead in order to motivate his view.Stokes writes: [We] have social infrastructure built around the dead that preserves them as objects of practical and moral regard.We have testamentary and probate laws, 5 Here, one may wonder if the norm, "Do not speak ill of the dead", is a genuine norm at all, since we are allowing for exceptions in our compliance.For example, we would not want to say that it is wrong to speak ill of Jefferson's ownership of slaves (Thanks to an anonymous reviewer for challenging us with this example).Our response is twofold.First, we think that this point helps to show that the norm in question should be understood to apply to The Dead as a group and not to particular members.This point, taken together with our lightweight treatment of norms (see footnote 1 above), would seem to mitigate the problem.Second (and more importantly), even if "Do not speak ill of the dead" turns out to be more robustly reason-giving in some sense, exceptions like Jefferson do not necessarily pose any problem.For, these cases-where our compliance with the norm is defeated by additional considerations-would at best lend to the conclusion that the reasons associated with the norm in question are pro tanto rather than definitive in nature.(See Broome et al., 2004Broome et al., , 2013 for discussions on pro tanto reasons) That said, the norm "Do not speak ill of the dead" should be understood to be defeasible, generally applying only in the absence of any obvious defeaters.

Synthese
(2022) 200:210 rules about disposing of remains and disturbing old burial plots, increasingly sensitive norms for the curation and display of human remains in museums, and material and ritual resources for commemorating the dead, from online condolence books to tombs of unknown soldiers.All of this ethical and social infrastructure serves to maintain at least something of the social identity and standing that the dead had before their demise[.](Stokes, 2019, pp. 761-762) Stokes takes the dead to be persons along the same spectrum of how we relate to those in permanently vegetative states (PVS) as persons.We commonly, and rightly, refer to them as persons.Moreover, while such cases may illustrate a severely degenerated sense of personhood, our social practices, norms, and laws nevertheless commonly consider such individuals as persons.Pushing this point a bit further, disallowing them their personhood status (which would already be controversial by itself) can easily incite a slippery slope when discussing the status of severely mentally disabled individuals.Rather than to draw a controversial line in the sand somewhere, Stokes advocates that the concept of person should function on a spectrum extending to degenerated cases including PVS individuals and even the dead.
Following Stokes, then, we can establish that the dead are persons by how our norms and social institutions substantiate this position.From here, we can argue that, because the dead are persons and because wide-spread external norms prescribing how we ought to engage with the dead exist, the dead constitute a social group-The Dead-just like PVS individuals constitute a social group.
One could deny Stokes' view by arguing that just because society happens to treat the dead as persons does not have to mean that they really are persons; society could simply be mistaken.But, if the dead are not persons then what are they?One could endorse (3) by saying that a dead person is not a person at all, but simply an inanimate entity.This entity would probably not be a concrete entity, since when we refer to Socrates, we do not refer to Socrates' lifeless body.The viability of (3) depends heavily on background assumptions.For instance, one could endorse permanentism.Permanentism is a view in metaphysics according to which everything always exists.Under this view, past and future entities are just as real as present entities.So, for instance, one could say that Socrates was a person between 470-399 BCE, and still exists as something that was a person between 470-399 BCE.A supporter of this view is Timothy Williamson who argues that Socrates exists as a non-concrete entity that once was a concrete person.(Williamson, 2013) If this view can be substantiated, then the dead would not be persons.Nevertheless, since we treat the dead as persons, they would still occupy "person-space"."Person-space" is a notion developed by Marya Schechtman (2014).For Schechtman, person-space is.
[T]he social and cultural infrastructure within which persons interact and which supports personhood.[…] The kinds of things that constitute social/cultural infrastructure might include religious traditions, institutions of punishment, codified systems of governance, economies, educational institutions, technologies, systems of symbolic representation through which information and knowledge are transmitted, means of transportation, and developed practices concerning arts, entertainment, and leisure.Of course, not every culture must contain all of these elements, and the details of such institutions will vary over time and place.(Schechtman, 2014, p. 115) Regardless of whether members of The Dead are persons or not, our treatment of them as persons substantiates their place in our social world, as having both linguistic and normative significance.If members of The Dead turn out to be degenerated cases of personhood, then their placement within person-space is rudimentary.And if, instead, the dead are not persons, they still remain within person-space, insofar as we keep relating ourselves to them normatively at a societal level. 78 The view we defend is thus that members of a social group need not be persons as long as they are treated as such by society.The dead need not be persons in order to be members of The Dead.However, this view would still presuppose that there exists something that can be a member of The Dead.Many would deny this presupposition and argue that the dead simply do not exist.Such a view seems intuitive, yet it is incompatible with a mainstream metaontological theory, which is Quinean metaontology.Quine explains the view via the slogan "to be is, purely and simply, to be the value of a variable" (Quine, 1948, p. 32).This means that the domain of the extension of the existential quantifier coincides with the domain of things that exist.In simple terms, if we quantify over something in our language, then such a thing exists.Our everyday talk extends to the dead-we can uncontroversially grant the dead properties, such as the property of having children-and therefore we treat the dead as values of variables when we refer to them.Therefore, under Quinean assumptions, we must grant that the dead exist.
7 By "treatment", we mean that we (1) talk of or to the dead as persons, (2) explain their lack of personspecific capacities by citing mitigating factors (i.e., that they cannot think, act, etc.only because they are dead), and (3) have norms and social institutions built around them being persons.Filling out "treatment" in this way suggests that norms surrounding some X must express a widespread social structure supporting the idea that X deserves to be treated as a person.It is not enough simply to point and say, "X, you are a person!".Requiring this larger social structure is useful in order to stem the potential overgeneration of external-norm-based social groups.This point explicates our point below that, in order to constitute a social group using external norms, it is useful to require that these external norms be sufficiently widespread.See also the following footnote. 8Some may protest that we are using Schechtman's notion of person-space incorrectly.First, because Schechtman's project is the identification of personhood while we are arguing that The Dead, as a social group, need not be populated with persons.Second, because Schechtman herself states that pets (and by extension non-agent objects) do not fall into person-space.(Schechtman, 2014, p. 122) Responding to the first objection, since Schechtman's position allows for cases of personhood without person-specific capacities (thinking, acting, etc.) (Schechtman, 2014, pp. 77-78), the external projection of personhood (by other persons) onto objects in these cases can become functionally equivalent to the objects being persons themselves.This point builds from Stokes's argument for extending personhood from individuals with PVS to the dead using Schechtman's framework.(Stokes, 2019, pp. 759-762) Moving to the second criticism, we can push back on two points.First, Schechtman allows for the possibility of non-human persons.(Schechtman ,2014, pp. 131-137) Second (and more importantly), Schechtman's denial of personhood in pets and objects (animate and inanimate alike) is not supported on metaphysical grounds.Rather, her denial of these cases is based on her disbelief that the widespread social and normative structures needed in order to externally imbue non-agents with personhood status is incredibly unlikely-society would need to genuinely embrace a kind of non-agent object (pets, stones, tables, etc.) as persons at the level of beliefs, norms, and social institutions.(Schechtman, 2014, pp. 136-137) In this way, Schechtman's view is not incompatible with ours.Rather, Schechtman's comments here make a good case for why our view does not easily lend to the overgeneration of external-norm-based social groups.

123
It is unsurprising that proponents of the view that the dead do not exist usually reject the Quinean framework, and accept rival frameworks, such as for instance a Meinongian framework, which allow for non-existent entities.A prominent example of a non-Quinean framework can be found in the work of Palle Yourgrau, who endorses (4), the idea that the dead are non-existent persons.In order to say that the dead are persons that do not exist, rather than saying that the dead are simply nothing, Yourgrau argues that non-existent entities can still be, and that existence is thus a substantive property that not all entities have.(Yourgrau, 1987, p. 89) According to Yourgrau (2019), most contemporary philosophers of death agree with him on the idea that the dead do not exist, and given the intuitive appeal of this view, we can call this the orthodox view on the metaphysics of the dead.It is thus of tantamount importance for the success of our account of the social group of The Dead that the existence of this social group is compatible with this orthodox view.Yet, there is an immediate problem which is that it is difficult to envision the existence of a social group which has members that do not exist.Fortunately, there are various ways out of this problem.
Yourgrau himself appeals to the difference between existing and being to solve this problem.Yourgrau speaks not of social groups, but of sets, but we take the solution to hold for social groups as well.Yourgrau says: It's absurd, I think, to suggest that when the elements of a set die, the set itself "dies" along with its members.Socrates and Plato, during their lifetimes, were obviously not "inside" the set {Socrates, Plato}, the way they were inside someone's home.Imagine if an old-timer, who knew Socrates and Plato, were to say to a young colleague, nostalgically, long after Socrates and Plato had died: "You young people today fail to impress me with your sets.You should have seen the sets that existed in my day."To my ears, at least, that sounds crazy.(Yourgrau, 2019, p. 83) Though Socrates and Plato no longer exist, they are still Socrates and Plato, and therefore they are still members of, for instance, the social group of Ancient Greek philosophers. 9The only special feature of the social group of The Dead would be that people would not remain a part of this group after dying but would become a part of the group when they die.Yet, we do not see any particular issues that would arise from this peculiarity.
Yourgrau's view is appealing in various ways, but it does rely on a distinction between existing and being which some might not want to be committed to.One philosopher who accepts non-existent entities, but who denies such a distinction is Niall Connolly (2011).Connolly instead supports (5), the view that members of The Dead are non-existing entities that are not persons.The reason for rejecting personhood is that 9 Yourgrau's distinction between existing as a narrow substantive property and being as something more general is somewhat reminiscent of Meinong's distinction between existing and subsisting.Meinong believes that all existing objects are spatiotemporal and concrete, whereas non-concrete objects could not exist but could only subsist.(Meinong, 1960(Meinong, [1904]], p. 79) If we maintain this Meinongian distinction, then social entities, such as universities, money, marriages, etc., could probably only subsist, not exist, since they are usually not taken to be concrete-they are not causally efficacious-or spatial.In this case, no social group would exist, but social groups would merely subsist.The existence or non-existence of their members would have no effect on their subsistence, and this would thus have no significant effect on the ontological status of social groups.
Connolly thinks that non-existent entities cannot possess any properties whatsoever.Instead, the dead, since they do not exist, can only be "bare particulars".According to Connolly, once a person dies, they do not merely lose their existence, but also their essential properties.10Connolly writes: Socrates doesn't have the qualities he had when he existed.He now stands in a relation to each of these qualities: the relation a thing stands in to the qualities it formerly instantiated.But he doesn't instantiate them.In particular Socrates doesn't instantiate being a philosopher.If he did he would still be a philosopher [.] […] Socrates is an object that has no qualities at all.Socrates has no features which serve to distinguish him, in and of himself, from the other denizens of the realm of non-existence.All the facts about Socrates are facts involving relations.These include the relations he stands in to the qualities he formerly instantiated.They also include the intentional relations he stands in to many living persons: the persons that love and admire Socrates.(Connolly, 2011, p. 95) While we speak of the deceased as persons, under the above framework they are not.As we understand Connolly, members of The Dead are quality-less objects that function in our linguistic and normative discourses via the relations they stand in both to us and to the qualities they once instantiated.And it is through these relations that members of The Dead, while nonexistent, have significance within our social world.
Connolly's view seems somewhat precarious to us, as it is not entirely clear how a non-existent bare particular could stand in relations to persons and how bare particulars can be distinguished using these relations, but on the condition that Connolly is correct and that there are indeed such relations between the social world and non-existent bare particulars, then these relations would thereby connect these bare particulars to their respective person-spaces in our social world.The strategy would be effectively the same as in the case of existing but inanimate entities: we show that the way in which society treats these entities (i.e., the norms, institutions, collective attitudes etc.) constitutes person-spaces for these entities-even though the entities themselves are not persons-and since there are these person-spaces, there are also the social groups to which these person-spaces belong.
However, there is the additional worry of the existence of the social group of The Dead, given that its members do not exist, and-as an anonymous reviewer pointed out-Connolly cannot appeal to Yourgrau's solution to this problem, because he does not make the distinction between existence and being.Nevertheless, we believe that, if Connolly's account of the relations between non-existent entities and persons can be made sense of, then this account will also show how the social group of The Dead can still exist even though the members do not exist.After all, it is clear that the person-spaces exist, but the question is how non-existing bare particulars are related to these person-spaces.It is these relations, argued for by Connolly, that are to offer us a bridge from the non-existent into the existent world.Again, this might seem precarious, but this precariousness is built into Connolly's account, and it should not warn against taking The Dead as a social group to exist, but it should warn against choosing Connolly's account over Yourgrau's account.Finally, one might reject both the idea that the members of The Dead exist and that there can be non-existent persons or objects.In this case, one would have to conclude that the dead are nothing.Such a view would have to overcome some hard issues about reference, and we are doubtful that this view could be made to work.However, in order to give a more complete overview of the possible metaphysical positions on the dead, we include it nevertheless as view (6).If the dead are nothing, then it simply does not make sense to speak of members of The Dead, and one would thus have to say that The Dead, as a social group, is memberless.The obvious question is: how could there be a memberless social group?We believe, however, that even in this case The Dead would be a social group.It would not be a social group if (6) were a widely endorsed view, because then society would stop paying homage to The Dead, treating The Dead with respect, having norms in place about not speaking ill or laws that prohibit building on burial grounds, etc.However, as long as this societal structure is in place, it constitutes person-spaces for the dead in the social world.According to ( 6), these person-spaces would remain empty, but this does not prevent them from constituting the social group of The Dead. 11  One might argue that for The Dead to be a social group there still needs to be something that we can treat as a person, i.e., something that our norms and institutions constituting person-spaces of the dead can apply to.Otherwise one could not even say something like "Society has constructed a person-space for my great-grandfather", since if the dead are nothing then there would be nothing a person-space could be constructed for.However, this problem is a part of the general problem of reference that supporters of (6) must face.Likewise, if the dead are nothing, one also could not say "My great-grandfather has great-grandchildren", since all ordinary discourse involving the dead would suffer from reference failure.If a solution can be found to this problem that would enable us to save our ordinary discourse and remain faithful to (6)-and we doubt that this can be done-then this solution would likewise enable us to say something like "Society has constructed a person-space for my great-grandfather".the thesis of methodological individualism.This thesis is described here by Katherine Ritchie: Methodological individualism is the view that social phenomena are to be explained and understood wholly in terms of their relations to the intentions and actions of individual actors, rather than any 'mysterious' social forces.If all social phenomena are to be explained solely in terms of individuals, social groups, one might argue, are explanatorily irrelevant and should be eliminated from one's ontology.(Ritchie, 2015, p. 311) Thus, according to methodological individualism, all features of social groups can be explained in terms of features of the individual members of these social groups.Ritchie believes that if methodological individualism is true, then social groups do not "really" exist, in the sense that we should not undertake an ontological commitment to them.The social group of The Dead presents a clear problem for this thesis, since it is unclear whether this social group contains individual persons at all, and even if it does, the normativity that attaches to the social group cannot be reduced to or explained by a normativity at the level of the individuals that make up this social group, since these persons would (probably) have no intentions (or any other mental states) and would not act at all. 12  The example of The Dead shows that one cannot reduce social groups to their individual members, because there would be a normative surplus.Whether there would also be a metaphysical surplus depends on how the normativity of social groups is to be explained.However, we take an explanation along metaphysical lines to be generally unfavorable.For instance, it is sadly still common in our society to give metaphysical explanations of the sexist norms which guide how we treat women, and to thereby use alleged facts about what women are in order to explain how our society treats women. 12We are indebted to an anonymous reviewer who interestingly suggested that an individualist about social groups may challenge our position as follows.Let N denote the norm "speak no ill of the dead".Now, the individualist claims, "N is not a norm concerning The Dead at all, but rather a norm concerning the interests of the living in a good posthumous reputation".The individualist thus recasts N as shorthand for a defeasible norm like N 1 : "for any person x, once x dies, do not besmirch x's reputation".We doubt whether all the norms concerning The Dead could be interpreted in such a way, since it is not immediately obvious that, for instance, the norm not to build on burial grounds concerns the living in the same way.More to the point, however, it is not obvious that N can actually be reduced (as shorthand) to N 1 .In order to do so, N 1 minimally needs to apply in all cases that N does.Now consider Ted, now deceased, who never had any interest in his reputation, nor in defending it.Let us also add that there are no additional facts about Ted's life that could defeat N (Ted was not a murderer, slave owner, etc.).Now, while N certainly applies (since Ted is dead), it is hard to see how N 1 could apply, since it is based on doing posthumous harm, and we cannot posthumously harm Ted (at least in this way) by besmirching a reputation he has no interest in.While most likely an outlier, Ted nevertheless does not seem unbelievable.The problem for individualists is that N 1 's application is contingent on the dead having posthumous interests in their reputations whereas N's application has no such requirement.Additional defeating considerations aside, speaking ill of the dead violates a social norm in the case of N but only possibly violates a social norm in the case of N 1 .Therefore, N being reducible to N 1 seems unlikely.As a final point, without content pertaining to individual interests, recasting N will not help the individualist.Since, recasting N as N 0 : "for any person x, once x dies, do not speak ill of x", for example, would still point to The Dead (albeit implicitly) in terms of membership criteria.In this way, the idea that norms for social groups can be stated without explicit mention of these groups is compatible with the view we develop later-that social groups are constituted by social norms.What it is to be a member of the elderly is to have a certain social status acquired by aging.Likewise, what it is to be a member of The Dead is to have a certain social status acquired by dying.This is clearly the opposite of what we should do, since this kind of explanation purports to give a metaphysical justification of why we ought to treat women in the way society does.Instead, we believe that there is no such justification at all, that women are treated badly in our society and are entirely undeserving of such treatment.According to Sarah-Jane Leslie, those who try to give such metaphysical explanations make the mistake of essentializing social groups.13She says: We essentialize a kind if we form the (tacit) belief that there is some hidden, nonobvious, and persistent property or underlying nature shared by members of that kind that causally grounds their common properties and dispositions.(Leslie, 2017, p. 405) We believe that this mistake of essentializing social groups is also made by Richard Miller, who argues that a kind is real only if the possession of it is intrinsic to its instances.For instance, he says, "copper is a real kind while being surrounded by sulphuric acid is not".(Miller, 2000, p. S649) He argues that the same must hold for social kinds, and thus also for social groups, because even if people are grouped together by external forces, they will not constitute a social group, unless they share in some kind of psychological identity.He deems this to be a necessary element of social groups, because it enables us to characterize "members' potential active causal roles in projects of mutual concern, resistance, and accommodation" (Idem, italics added).
This necessary condition presupposes that in order for something to be real, it must be causally efficacious, for Miller's only reason for citing the shared psychological identity is to account for social groups in terms of their causal roles.Miller seems to imply also that either something is real or it does not exist.In other words, Miller believes that we should either be naturalists or eliminativists about social kinds (where naturalism means that social kinds exist mind-independently and eliminativism means that social kinds do not exist at all).14However, we believe that it is better to contrast real with constructed; either something is real and can be found in nature or it is constructed and does not exist outside of our minds, i.e., its existence is mind-dependent.Social kinds are widely believed to be constructed (e.g., Searle, 1995;Hacking, 1999;Ásta, 2018).For example, money is constructed, since it does not exist without there being humans who conceive of its existence.If one believes that The Dead is a social group, then the members' possession of a shared psychological identity cannot be a necessary condition for the existence of social groups.Therefore, since we believe that The Dead is indeed a genuine social group, we must reject Miller's causal-reductionist condition and the naturalism it implies.Instead, we endorse constructivism about social groups.
Constructivists about social entities believe that these entities are the products of our practices, attitudes, beliefs, linguistic acts, and so on.Our constructivist approach to social groups is to claim not just, like Thomasson does, that we can use social norms to individuate social groups, but that, ontologically, social groups are nothing over and above social norms.For this view, we can draw inspiration from Wilfrid Sellars' paper "Abstract Entities", in which he argues for a theory of universals that avoids both Platonism and nominalism.Very crudely construed, he argues that a universal such as triangularity cannot justify or explain the correctness of our application of our concept of triangularity.Rather, the rules of application are intrinsic to the concept of triangularity; they are conceptual norms.Triangularity then only exists as a reification of these conceptual norms.(Sellars, 1963) If Sellars is correct, then universals do exist and cannot be reduced to their instances, but their existence is constructed out of the norms that guide our linguistic practices.(Kraut, 2016).
We do not endorse this view on universals, but we do think it allows for an interesting parallel with social groups.Above, we rejected the essentializing move of saying that the social norms of a social group are to be explained or justified in terms of certain essential characteristics of the members that belong to the social group.We argued that in certain cases it would be misguided to look for a justification at all.Perhaps, then, instead of appealing to the nature or essence of a social group to explain their related social norms, we must explain the social groups in terms of their related social norms.Social groups would then exist as reifications of social norms; norms that we project onto people (and other objects).In this case, social groups are nothing over and above the relevant social norms.If we buy into this conception of social groups, then, just like universals on Sellars' account, social groups do exist and cannot be reduced to their instances (their members), but their existence is constructed out of the norms that guide our social practices.We can draw the example with money.Money is not a concrete entity.Coins and paper can represent money, but money cannot be reduced to coins and pieces of paper.Money is an inherently normative entity.Paper bills started out as explicit promissory notes that reflected that one person owes another person.Instead of ontologically reducing money to pieces of paper, we could ontologically reduce it to all the moneyrelated interactions and the rules (the norms) that facilitate these interactions.In this sense, money is a constructed entity that was constructed in order to keep track of commitments and entitlements.The entity itself can be thought of as a reification of these normative relations.Likewise, social groups can be thought of as reifications of social norms, since they do not amount, ontologically, to anything more than our social practices and the norms that facilitate these practices.

The dead versus social group Internalism
Another curious feature of the social group of The Dead is that there could not possibly be any internal norms, since the members of the group are incapable of acting (and are thus incapable of possessing norms as to how to act).The social group of The Dead challenges the view we call Social Group Internalism, according to which the presence of certain elements internal to a collection of people is essential for the creation and existence of a social group.Social Group Internalism does not regard external norms as particularly relevant to the existence of social groups.Though internal norms might generate external norms, or external norms might even generate internal normativity, internal norms are seen as an essential component for social groups.Yet, if The Dead is a social group, then it is so solely in virtue of the presence of external norms.While Social Group Internalism is a widely disputed view, there are definitely philosophers who support it.For example, by appealing to the special notion of "We*", Margaret Gilbert argues that a social group is a collection of people that has come together to form a plural subject.(Gilbert, 1992, p. 204) Without going into detail about her account, and by putting things very roughly, a social group is a collection of people that have collectively accepted to do things together as a "we" or "us".Gilbert's view counts as internalist given that what is necessary and sufficient for the creation of a social group is that plural subjecthood, a condition internal to the relevant collection of people, is obtained amongst the members of that collection.
Another example of Social Group Internalism is Miller's account of social groups mentioned in the previous section, according to which things like shared identities, common concerns, and mutual goals as a collective, are what makes the collection a real social group.(Miller, 2000, p. S649).
Though Thomasson does not think of the presence of internal norms as necessary for social groups, there are some compelling reasons for why they would be.Ritchie argues that there are two types of social groups; the first type includes those groups that are structured, such as committees, whereas the second type includes the unstructured social groups, such as genders, ethnic groups, groups based (solely) on sexual orientation, etc. (Ritchie, 2015) Internal norms seem clearly indispensable to the former type of social groups, but the second type of groups seem to be primarily characterized by external norms.However, even for this kind of group, it seems impossible that there are not also internal norms, since if a group has external norms, then, provided that the members are, to a certain extent, aware of these norms, these norms will inform their actions and thus create derivative internal norms.For example, members of a marginalized group are often very cautious in their dealings with people outside of this group.If it were the case that members of a marginalized group are completely oblivious of the negative stigmas surrounding this group, and do not let their membership of this group guide their acting in any way whatsoever, then they would perhaps not think of themselves as members of this group at all, nor would they think that such a group exists.It is true of course that sometimes the members of marginalized groups fail to realize that they are being marginalized, but while they might fail to realize this, they do let themselves be guided by the internal norms of the group.Therefore, even for the second type of social groups, the presence of internal norms seems to be ubiquitous.Thus, the fact that there are social groups that lack internal norms, which is evidenced by our example of The Dead, might complicate matters for normativity-based accounts.
One potential complication would be that if we drop the presence of internal norms as a necessary condition for the existence of social groups, and thus claim that external norms alone are sufficient for the existence of social groups, then our normativity-based account of social groups would be too permissive, because it would count too many assemblages of people as social groups.We could call this the Too Many Groups problem.From the above, we can see that it is tempting to think of internal norms as necessary for social groups, since it is difficult to think of instances of social groups that lack any internal norms whatsoever.However, this argument cuts both ways.We showed that whenever there are widespread external norms, there will most likely also be internal norms, so it would be very difficult to think of counterexamples of collections of people that are solely constrained by external norms.The only instances in which this would not happen is when the members of a social group are unable to recognize these external norms, which would be the case only in very limited circumstances, e.g. when the members of the social group are all dead.We are already committed to these limited cases anyways and we believe that we already built a good case for why they count as genuine social groups.Therefore, it seems that the problem of Too Many Groups does not arise on account of the presence of internal norms not being a necessary condition.
If anything, the Too Many Groups problem seems to arise when we claim that the presence of internal norms is sufficient for the existence of social groups.Consider the following case of a collection of people accidentally sharing internal norms.Pick some random property, like having a pinky toe that is x cm long.Imagine that there are precisely 10 people in the world who have this specific length of a pinky toe, and that they all live in different parts of the world.Each believes that there will be some number of people other than themselves that share this particular property, but they have never met each other and perhaps they never will.By some miracle, each member of this collection comes up with similar sets of norms regarding how they, the special pinky length people, ought to behave and each of them follow these norms.It is hard to conceive this collection of people as a social group.It is not likely that they ever come together and do things together.There are no external norms concerning how people treat these special pinky toe people.It is just that some collection of people sharing some specific property accidently happened to internalize similar norms and each acts on them individually and independently from each other.Of course, these individuals could try to create a genuine group out of this shared arbitrary property, but this would require the presence of some norms that are not internal to the individuals, such as structuring norms (which are internal to the group but are external to the individual members) or external norms.They could do this by getting in touch with potential members of the group and creating a network or by spreading the word to force people outside of the group to recognize the group's existence.These strategies further corroborate the idea that internal norms are not sufficient by themselves.
Therefore, abandoning the idea that internal norms are necessary for social groups does not create the Too Many Groups problem, but abandoning the idea that internal norms are sufficient does help us fix the Too Many Groups problem.One could only conclude from this that we have been too preoccupied with internal norms and not enough with structuring or external norms.After all, the example of The Dead shows us that, while internal norms may not be sufficient for the existence of social groups, external norms are.
Another potential concern for our view is that, by arguing that external norms are sufficient for social groups whereas internal norms are not-and by thereby showing external norms to be, in a sense, more important-we might be committing ourselves to some unfortunate and problematic views.For instance, we would want to say that women are not to be defined by how men think about them.However, it might appear that we cannot say such a thing if we argue that external norms are more important than internal norms.Yet, we do not believe that this is a problem for our view, but rather, it is a problem for society.The sad reality is that, in our current society, women (as a social group) are largely defined by how men think about them, in the sense that a patriarchal society creates patriarchal norms and divides up the social world accordingly, such that the very concept of woman is a man's invention.

Conclusion
We have used Thomasson's normativity-based criterion for social groups to argue that we must think of The Dead as a social group.We must do so because the norms we have that guide the ways in which we must treat The Dead, such as "De mortuis nil nisi bonum", indicate that we do in fact take The Dead to be a social group in many if not all societies.Taking The Dead to be a social group led us to accept that not just persons, but also other objects, can be members of a social group.From this, we developed an account of social groups that takes social groups to be reified social norms.There are some important implications to accepting that The Dead is a social group.We have discussed two such implications.The first is that it allows us to argue powerfully against individualism about social groups.This is because it is clear that the social group of The Dead, and its related norms, cannot be reduced to its individual members, since the group exists regardless of whether or not its members exist.Another implication is that the presence of internal norms is not necessary (nor sufficient) for social groups.Clearly, the social group of The Dead does not have internal norms, because there are no ways in which its members can act, and thus no ways in which they can be guided to act.We have argued that none of these implications suggest that we must not treat The Dead as a social group, and we believe that they actually help us get clearer on what social groups are.
2. Members of The Dead are inactive persons.3. Members of The Dead are inanimate entities.4. Members of The Dead are non-existent persons. 5. Members of The Dead are non-existent entities.6.The Dead does not have any members.