Political Science, St. Louis University, St.
Louis, Missouri, USA
Correspondence
Ruth Porter Groff, Saint Louis University,
St. Louis, MO, USA.
Email: [email protected]
Abstract
It is often assumed that social and political thought have
nothing to do with issues of concern to metaphysicians.
I have referred to this assumption in the past as The
Myth of Metaphysical Neutrality. I argue here that social and political theories have metaphysical positions
built into them, such that to adopt a given social or political account commits one to that theory’s implicit metaphysics — and, conversely, that commitment to a given
metaphysical position will preclude adopting social or
political accounts that are at odds with it. I look first at
the issue of emergence, showing that key concepts employed by Aristotle, Rousseau and Marx, respectively,
require a belief that wholes do not reduce to their parts.
I then turn to Marx’s account of alienation, arguing that
it presupposes a belief in agent-causal free will.
K E Y W O R D S
Aristotle, critical theory, Marx, social metaphysics, social
ontology
O R I G I N A L A R T I C L E
The devil is in the categories: Metaphysics and
social and political thought
Ruth Porter Groff
DOI: 10.1111/jtsb.12320

1 | INTRODUCTION
A common assumption in contemporary Anglo-analytic philosophy, including in social and political philosophy, is that whatever it is that we end up saying about the world of so-called ‘middle-sized dry goods’ (including us), it will be equally compatible with any underlying metaphysics. Social scientists, too, often presume that any true claim about what the social world is like
can be underwritten by any ontology, such that those who seek to explain social phenomena have
no particular stake in how questions of basic metaphysics are resolved. I have elsewhere referred
to this belief as ‘the myth of metaphysical neutrality’ (Groff, 2012). The myth of metaphysical
neutrality may seem plausible at first glance, but it does not hold up well under scrutiny. If one

looks to the history of Western social/political thought, one immediately finds examples of concepts, claims and distinctions that have been advanced that are not, in fact, consistent with any
and all metaphysical frameworks. Perhaps ironically, evidence for the persistent hold of the myth
comes not so much in the form of a raging or extensive debate in the literature, but rather in its
being ubiquitous enough as a background assumption that it apparently need not be vociferously
defended in print.
This said, let me begin with a caveat. It is obviously possible to act in ways that are at odds
with one’s theoretical convictions – or, for that matter, for one’s actions not to be theoretically informed at all, at least not explicitly. It is also possible to advance claims about the nature of social/
political phenomena without caring what the metaphysical prerequisites are for those claims
having the content that one understands them to have. These points I readily concede. Whether
we like it or not, however, our underlying conceptual categories do establish parameters within
which our more substantive beliefs may fall (at least if our views are coherent), and what we
believe does often (even if it need not necessarily) have a bearing upon how we act.1 It therefore behooves us, I think, to attend to our most basic philosophical assumptions, including the
metaphysical ones, even if there is nothing to stop us from ignoring what we find when we do.
For those readers – e.g., critical realists – who may already recognize the myth of metaphysical
neutrality to be false, I hope that the debunking of it that I offer here will be of interest even so.
If nothing else, my examples may provide new ammunition for those concerned to defend, formally or informally, their own belief that underlying metaphysical assumptions have a bearing
upon theories of social phenomena.
The discussion to come is organized into two parts. I look first at the dispute between ontological individualists and those who affirm the existence of irreducibly macro-level social entities, relations and properties. I then turn to the concept of alienation. In both cases, I argue, the
manner in which pertinent metaphysical questions are answered makes a difference for what
can be said about the relations of power in which we middle-sized dry goods find ourselves. The
structure of the reasoning in Part 1 is: “If you accept metaphysical principle mp, you may not
employ social/political concepts that are predicated upon mp being false.” In section two, the order is reversed: I start with a well-known social/political concept, and show what it presupposes
metaphysically.
2 | MACRO-LEVEL RELATIONAL ENTITIES AND SOCIAL
PROPERTIES
Let me begin with the issue of emergence. The contemporary literature contains a plethora of
definitions of the term, and I do not mean to argue about which one is correct, or about whether
or not emergent phenomena exist. My intent, rather, is to show that in the history of Western
social/political thought there are concepts, and lines of argument into which they fit, that may
be adopted only if one thinks that emergent phenomena (as per the definition that I shall utilize) do exist. As I shall use the term, emergent phenomena include the following: (a) relational
wholes, the being of which is not exhausted by the being of their constituent parts or relata; and/
or (b) properties of such wholes; and/or (c) properties derived from, grounded in, anchored by
or otherwise contingent upon the existence of an emergent whole or a relation of one kind or
another.2 With respect to the social world, (a) refers, paradigmatically, to corporate or relational
entities that are composed of individuals, and/or whose relata are individuals, but which, per
stipulation, are not metaphysically equivalent to a plurality of individuals (examples include a

family, a bargaining unit, a corporation, a society; and perhaps the relationship of employer/
employee, though one might believe this to be a property); (b) refers to properties of such entities
(e.g., a regime type; a characteristic distribution of power and/or resources); (c) refers to properties that are borne by individuals but that are, in one way or another, derived from properties
of a whole (e.g., a class, racial &/or gender identity; being an American citizen). Language(s),
traditions, norms and other inter-subjectively held and enacted symbols and meanings I will also
count as emergent; a fourth component may be added to the definition that I have given, so as to
accommodate such phenomena, or they may be counted as falling into one or more of the other
categories. Notice that one may believe in the existence of an emergent social entity that is itself
composed not of individuals, but of groups, relations or roles. Roy Bhaskar, for example, understood society to be such an entity – an emergent whole composed of relations (Bhaskar, 1979,
1998). Similarly, the Aristotelian household looks to be composed of the marriage relation, the
parental relation and the master-slave relation.
Those who deny the existence of emergent social phenomena, maintaining instead that it is
only propertied individuals who exist, have traditionally been called either ‘atomists’ or ‘ontological individualists,’ depending upon disciplinary context. I use these terms interchangeably,
since they are often equated and some readers may be more familiar with one or the other label.
Technically, one might want to reserve the term ‘atomist’ for the more general claim that it is
only part-less simples that exist. One might think that emergent social phenomena do not exist
(and that individuals do), and yet not believe individuals to be part-less simples – which position,
assuming a more exacting use of the terminology, would render one an ontological individualist
but not an atomist. Conversely, one might believe that macro-level social phenomena do exist,
but that it is a condition of their bona fide, irreducible reality that they are part-less, in which
case, if one were using the term ‘atomism’ in the stricter sense, one would count as an atomist
but not as an ontological individualist. Tim O’Connor and Jonathan Jacobs have defended such
a view of emergence (i.e., one in which emergent phenomena must be part-less) in the case of
individual agents, whom they take to be emergent vis-à-vis the physical components of their
bodies (O’Connor & Jacobs, 2003). Philosophers are more likely to use the term ‘atomism’ to refer
to the view that all existent entities are simples than are social/political theorists. The concept of
emergence as I have defined it does not commit one to the idea that emergent social phenomena
have no parts – or, for that matter, that there is anything especially mysterious about them epistemologically. My own view is that emergent sociological wholes do have parts, and that they can
act upon their parts, but nothing in the way that I will be using the term emergence precludes
one from holding either the view that emergent phenomena exist but are by definition part-less
or that they exist but do not act upon their parts.
Durkheim referred to emergent sociological phenomena as ‘social facts,’ distinguishing them
from ‘psychological facts’ (Durkheim, 1964). (In The Rules of Sociological Method, Durkheim
does not differentiate sharply between phenomena and statements about phenomena, but the
ontological sense of the term ‘social fact’ undergirds its epistemic sense.) Sociology, he declared,
differs from psychology in that its object domain is the former rather than the latter. John Stuart
Mill, by contrast, denied the existence of such phenomena, which he characterized as involving
what he called ‘chemical’ rather than ‘mechanical’ composition (though as it happens he did
allow for emergent psychological phenomena, vis-à-vis biological, or physical phenomena, as
well as for emergent molecular phenomena) (Mill, 2006). Societies, Mill maintained, are not
akin to water molecules (his example). Hydrogen and oxygen “duly arranged,” to use the contemporary parlance, constitute a new propertied thing; individuals duly arranged do not. Contemporary thinkers have introduced into the discursive mix various intermediate positions (and

accompanying terms, e.g., “weak” versus “strong” emergence), meant to do justice, conceptually,
to ranges of commitment relative to different definitions of emergence, be those definitions ontological (as is the one that I have employed) or epistemic, or sometimes to distinguish between
epistemic and ontological definitions. For present purposes, such maneuvers are likely to be distractions. What is of interest is the distinction between affirming or denying the existence of
social facts (construed as macro-level sociological phenomena that do not reduce to individuals
or to properties thereof) in addition to the existence of psychological facts. My claim, which may
appear to some readers to be self-evident, is that there are things that may be said – things that
have been said – that may be properly said only if one believes that in addition to it being the case
that individuals exist, it is the case that emergent sociological phenomena exist.
Let us begin with irreducibly social macro-level entities (i.e., propertied things), then turn to
irreducibly social properties (though the distinction is an analytic abstraction; in reality there
are neither property-less things nor free-floating properties). A ready example of an emergent social entity, for those who are familiar with the history of Western social and political philosophy,
is the polis as conceived by Aristotle in the Politics (Aristotle, 1998). The polis, which Aristotle
distinguishes from what he dubs a “mere alliance,” is a multi-party relationship between citizens,
where ‘citizen’ is defined as one who takes part in the administration of justice, as Aristotle puts
it. Aristotle describes the polis as being natural, and as existing “prior” to the individual. It is the
highest form of association, he tells us, inasmuch as its proper purpose is to be both the expression of, and the venue for, human flourishing. Someone who denies the existence of irreducibly
social macro-level entities will have to say about such a purported entity: “Look, I grant you
that Aristotle believed in the existence of poleis, but the metaphysics was bad. There is no such
thing as a polis. There are only individuals.” If, for some reason, such an interlocutor wanted to
re-purpose Aristotle’s language (despite the confusion that such a step would predictably cause),
such that she too could make realist claims about something that she too would call a polis, she
would have to assign a new referent to the term, stipulating that when she uses it, the word
‘polis’ does not refer to what it refers to when Aristotle uses it. (In the way of such maneuvers,
the refurbished noun could be written as polis*, to mark the change.) She might even be able to
translate some of Aristotle’s statements about poleis into statements about individuals (i.e., about
poleis*), at least well enough for the new claims to be intelligible. But she would not thereby be
talking what Aristotle had been talking about. On the contrary, she would be talking about the
properties and purposes of individuals, whereas Aristotle had been talking about the properties
and purposes of irreducibly relational entities of the specified kind, the existence of which he
affirmed and she denies.
A certain style of ontological individualist might contend that Aristotle himself was talking
about poleis* rather than poleis, even though he did not think that he was, since (in her view) it
turns out that some number of individuals greater than one is what a so-called polis actually is.
Such a response might be dismissed as so much sophistry, but even if it were to be permitted, the
atomist has not gained any ground by it. She has simply reiterated her own position. If the thesis
of ontological individualism is true, as the ontological individualist believes it to be, then Aristotle was mistaken to affirm the existence of poleis. It does not follow that Aristotle himself, in theorizing the nature of the polis, was actually theorizing what it is to for there to be some number
of individuals. Those who accept ontological individualism and those who reject it disagree, fundamentally, about how the world is; whatever our ontological individualist’s theory of reference
may be, it will have to allow for the fact of such disagreement. Aristotle was talking about poleis,
not poleis.* And besides, the thesis of ontological individualism might very well not be true.

For better or for worse, Aristotle himself believed in the existence of entities such as poleis,
the nature of which he was therefore in a position to theorize. As I have just observed, some
of what Aristotle tells us about poleis one might be able to say about individuals. For example,
Aristotle reports that the proper purpose of the polis is to facilitate bona fide human flourishing,
and even he holds also that there are individuals who recognize that bona fide human flourishing ought to be their aim, and who act accordingly. Someone who believes in the existence of
individuals but not in the existence of emergent political associations could presumably talk
about the distinctive traits of wise individuals, including what such persons take the natural
telos of a human life to be, even if she could not tell us anything about the proper purpose of a
polis – and therefore could not give us Aristotle’s own account of the phronimos. Similarly, an
ontological individualist could say that multiple individuals are “prior” to any one individual (at
a minimum you need at least two individuals in order to get a new one), and/or that individuals
are natural entities. Once again, however, the ontological individualist would not be telling us
anything about Aristotelian poleis – let alone if, in addition to denying the existence of emergent
phenomena, she were also to deny the existence of essences, since it is our form that connects
us necessarily to the polis, as Aristotle has it, rendering us political and the polis natural. But at
least there would be an intelligible claim to be made, even if, as per the ontological individualist’s
ontology, it is a claim about individuals, and not about poleis.
That an ontological individualist can apply to individuals some of the descriptors that Aristotle applies to poleis is a far cry, however, from her being able to say what Aristotle said about
the polis. Moreover, other aspects of Aristotle’s thinking do not allow for even this much, in the
way of appropriation. For example, Aristotle thought it necessary to talk about the telos of a polis
in addition to establishing the purpose or function of individuals. The second step would be redundant for an ontological individualist, since poleis* just are individuals. The ontological individualist might reply by saying: “Well, a polis* is individuals, but it is individuals duly arranged;
that’s why we have to ask the telos question a second time.” But this would be an equivocation.
Anyone who believes that entities such as poleis exist will agree that their parts are duly arranged
– or at least anyone will who admits of the existence of entities that have parts. The question
is whether or not the arrangement amounts to anything, metaphysically. Those who believe in
what Mill called chemical composition think that hydrogen and oxygen, duly arranged, constitute something that is not simply a quantity of hydrogen and a quantity of oxygen. Whether or
not this type of composition occurs at the level of the social is precisely what is in dispute. And it
will not do for the ontological individualist to try to have it both ways. If the stipulation of ‘duly
arranged’ is meant to indicate the formation of a new entity, not equivalent to the being of the
plurality of individuals that compose it, then our putative ontological individualist is not an ontological individualist after all. If that is not what the stipulation is meant to indicate, then nothing
has been said to render Aristotle’s need to pose the second question any less opaque, from an ontological individualist perspective, than it was before. Notice that Aristotle also asks about what
makes for the happiness of a polis, and not just about the telos of poleis.
Similarly, after telling us that “when the constitution of a city undergoes a change in form …
the city will likewise cease to be the same city,” [1276b1] Aristotle illustrates the point by saying
that a chorus that changes from comic to tragic is no longer the same chorus, even if the people who are in it do not change. The claim presupposes that there is an ontological gap, to use
Roy Bhaskar’s term, between the emergent entity, here a chorus, and the individuals of which
it is composed (Bhaskar, 1978, 1979). We are more accustomed to hearing about corporate entities or even structural relations that remain unaltered despite a change of personnel, rather
than the other way around, but the ontological gap is the same. Or again, when, reflecting upon

democracies, Aristotle allows that the many who are unwise might in fact exceed in wisdom the
few who are wise, he notes that he means the many “collectively and as a body,” adding that:
“when all meet together, the people may thus become something like a single person.” [1281a39]
Unlike Hobbes, Aristotle does not liken the polis itself to a single individual human being. He
does, however – as already noted – distinguish between a polis and a “mere alliance,” ascribing of
a degree of unity and integration to the former that he does not believe is had by the latter. Poleis
are genuine collectivities; alliances are not. Moreover, as we learn from the account in Book 1 of
the relative “gregariousness” of human beings compared to that of bees, poleis are collectivities
the nature and purpose of which is tied to the fact that human beings use language (Greek specifically) to deliberate about good and bad, a phenomenon that is itself difficult to render as an
individual, psychological fact rather than a social fact, as per Durkheim’s classic discussion of the
matter (Durkheim, 1964).
Of course, Aristotle is not the only prominent political philosopher who assumes the existence of irreducibly corporate entities. Consider Rousseau (if not Hobbes). In the Social Contract
we are told repeatedly, and in no uncertain terms, that a Sovereign is a singular entity, not simply
multiple individual persons (Rousseau, 2003). Qua unified whole, the Sovereign is Rousseau’s
proposed solution to the crisis of political legitimacy that arises in societies that actually are the
way that the ontological individualist imagines all societies to be. Rousseau’s question – “How
does a people become a people?” – cannot even be posed, let alone given Rousseau’s answer, by
someone who believes, as a matter of ontology, that ‘a people’ is metaphysically equivalent to
‘multiple persons.’ To be clear, the ontological individualist can ask how any number of individuals might come to hold a given belief, and/or how individuals might come to act in one way or
another. But that is not what Rousseau is asking when he asks how an aggregate of persons, each
with their own particular will, comes to be a Sovereign, with a General Will. Indeed, anticipating
later thinkers who draw upon his work, Rousseau already prompts us to wonder if the principled
rejection of the existence of anything but particular wills might not itself be an indicator, at the
level of thought, of a society marked by the very fragmentation and alienation that Rousseau
sought to remedy. It is worth noting further that the Aristotelian polis is said to have parts, whereas the Rousseau-style Sovereign most certainly does not, and also that which model one may
adopt (assuming that one affirms the existence of emergent entities) will itself be constrained by
one’s metaphysics. If one believes in the reality of social wholes that do not reduce to individuals,
but believes that such wholes must be part-less, then one will have to opt for Rousseau’s model
rather than Aristotle’s.
Another example of an irreducibly social entity that one will not be in a position to include
in one’s social/political theory if one denies the existence of emergent social wholes is the actor
that Marx calls the ‘collective worker’ or ‘collective laborer,’ introduced first in chapter 13 of
Capital, Vol. 1, the chapter on the dynamics of co-operation in the context of capitalist industrial
production (Marx, 1992). It is worth quoting Marx at some length here (although this is just one
of several passages in which he describes the collective worker in the following terms):
Just as the offensive power of a squadron of cavalry, or the defensive powers of an infantry
regiment, is essentially different from the sum of the offensive or defensive powers of the individual soldiers taken separately, so the sum total of the mechanical forces exerted by isolated
workers differs from the social force that is developed when many hands co-operate in the same
undivided operation, such as raising a heavy weight, turning a winch or getting an obstacle out
of the way. … Not only do we have here an increase in the productive power of the individual,
by means of co-operation, but the creation of a new productive power, which is intrinsically a
collective one (p. 443).

Marx points out further (adding another layer of irreducibly sociological content to the analysis) that this “unification [of workers] into one single productive body” is not “their own act (p.
449).” Rather, it is a function of their having been brought together by capital – for which reason,
he says, the power of the collective laborer seems to be a power borne by capital, “appear[ing] as
a power which capital possesses by its nature – a productive power inherent in capital (p. 451).”
The claim that capitalism itself gives rise to an emergent, class-based collective subject is not
simply Marx waxing metaphorical. Apart from the political import of the notion of a class in
itself, the concept of the collective worker is directly connected to the distinction between formal
and real subsumption, and to what Marx has to say about the social form of specifically capitalist
laboring. “The concept of a productive worker,” he writes in chapter 16, “… implies not merely
a relation between … the worker and the product of his work, but also a specifically social relation of production, a relation with a historical origin which stamps the worker as capital’s direct
means of valorization (p. 644).” As before, with respect to the Aristotelian polis and the Rousseauean Sovereign, my point is not that the ontological individualist is precluded from making claims
about the behaviors, beliefs or attitudes of multiple individuals (though, as I also noted before, it
may well be that even her claims about individuals cannot hold up, e.g., if it turns out that there
is no satisfactory account of language according to which it is something that reduces, metaphysically, to the mental states of individuals). Nor am I insisting that Marx, in this case, was correct
in his analysis, though he may well have been. Here too, the point is simply that there are claims
that Marx made that one cannot make if one disavows the ontology that they presuppose. As in
the Aristotle case, the ontological individualist may be tempted to assert that really Marx did not
mean what he expressly says that he meant; really, Marx too was an ontological individualist.
Interestingly, this move is less commonly attempted in relation to Rousseau. In practice, ontological individualists are more likely to flatly reject the existence of an entity such as Rousseau’s
Sovereign, with its General Will, than to claim that really what Rousseau had in mind was not a
collective subject but a numerical plurality of individuals.
There is nothing particularly abstruse about what I have been saying. Manifestly, someone
who holds that emergent social entities do not exist is not well-positioned to advance claims the
sense of which depends upon the existence of such entities. Granted, there is a concept, viz.,
“non-eliminativism,” which is called upon precisely for the purpose of allowing one to talk as
though something the existence of which one denies, actually exists (“I don’t believe that x’s exist, but I’m not an eliminativist about x’s,” one says, if one thinks that there are no such things
as x’s, but nevertheless wants to be able to refer to them in an ostensibly realist fashion), but a
declaration of non-eliminativism does not suspend the law of non-contradiction. It is the same
with respect to emergent properties (which, again, in reality do not float free from the propertied
things whose properties they are) as with entities. To begin with Rousseau this time, the General
Will is construed as an emergent feature of the Sovereign. Rousseau specifically reminds us that
it is not to be confused with the particular wills of individuals. The ontological individualist will
not be able to avail herself, conceptually, of such a property. Nor will she share, or even be able
to share, Rousseau’s own concern with the ever-present risk of the General Will being displaced
by a plurality of particular ones. Similarly, with respect to the Politics, since being a given regime-type – an aristocracy, say – is, as Aristotle has it, a property of poleis, and not of individuals,
it is not a property to which one who denies the existence of social wholes such as poleis may realistically refer. Being a citizen is a property of individuals, but the ontological individualist will be
no better able to engage in Aristotelian talk about aristocratic citizens than she is to talk about
aristocratic poleis, since – for Aristotle – citizenship of any type is a function of being a member

of a polis. Being a citizen is a property borne by individuals, but it is an emergent, social property
of individuals, a social fact rather than a psychological, or individual fact.
It will be the same with any of the emergent properties that figure so centrally in Marx’s work.
The General Formula for Capital, for instance – M-C-M’ – straight-forwardly names a complex
property of capitalist societies (in keeping with Aristotle, Marx understands societies to be internally differentiated relational wholes, entities that are not equivalent to individuals and the properties thereof). A more theoretically complicated example is the phenomenon of value. Marx
himself talks about value being what he calls a ‘social substance’ rather than a property, but for
present purposes it does not matter whether we think of it as a property or a (social) substance.
I am treating it as a property since what a quantity of value is, is a portion of the total alienated
collective capacities-to-do of a capitalist society, but regardless of how we classify it, value as conceptualized by Marx is an irreducibly social phenomenon. Whether he is talking about the value
that is associated with a given duration of expenditure of commodified labor-power or the value
had by some other commodity, a commodity’s having (or being) a value at all is akin to a person
speaking a language: it presupposes not just multiple individuals accidentally (and miraculously)
doing the same things for the same reasons (or using the same sounds in the same ways), as per
the ontological individualist, but the existence of an organized whole, here an economic system
of a particular (and peculiar) kind. Marx might have had it wrong, of course. But if one wants
Marx’s concepts, one has to accept the ontology that comes with them. The General Formula for
Capital and value are examples of emergent properties of capitalist society as a whole. At the level of individuals, just one example in Marx’s work of an emergent property borne by individuals,
akin to the example of Aristotelian citizenship, is the property of being a capitalist (or a wage-laborer). Indeed, the fact that (as Marx has it) the property of being a capitalist is a function of the
class character of the social whole – this fact has led some Marxists to argue that Marx denies the
very existence of individuals.
I do not want to belabor the point (no pun intended). If, as a matter of ontology, one rejects
the existence of emergent social phenomena, then, having denied their existence, one will not be
able to help oneself to positions or concepts at the level of social/political analysis or explanation
that presuppose their existence. Conversely, if one asserts, for example, that there is a difference
between structural racism and racist acts by, and/or beliefs of, individuals, then one cannot also
maintain – and patently one does not actually believe – that the former just is the latter, since the
distinction depends upon the phenomena not being equivalent.
3 | ALIENATION
I want to turn now to the posited phenomenon of alienation, as conceptualized by Marx. As an
epistemic matter, the concept of alienation refers to the mistaken belief that one’s own efficacy
is the efficacy of something other than oneself. Feuerbach had said that what we imagine to be
God, is actually us. Marx applies that idea to society: it is the same with market forces as with
God, he says, in that we do not realize that “the numbers,” for example – to invoke Kai Ryssdal of NPR’s “Marketplace” – are actually us. But Marx goes further. Unlike God, Marx thinks,
market forces are real. They really do constrain our options for acting. It is in virtue of this that
alienation, as Marx has it, is not just a subjective condition of error, but an objective condition in
which our agency has been genuinely curtailed. Moreover, the powers that we unknowingly turn
against ourselves when we are in such a condition are not just any old powers. That which we
both correctly and incorrectly take to be an external, coercive source of determination is nothing

other than our distinctively human capacity for creative self-determination, our ‘species-being,’
as Marx puts it in 1844, distorted and become alien to us (Marx 1964). This aspect of the situation renders alienation not just an objective condition, but a morally bad objective condition. I
am not concerned here with whether or not Marx’s analysis is correct. My claim is only that it
is not consistent with any and all underlying metaphysical frameworks. On the contrary: Marx’s
concept of alienation presupposes the existence of agents with free will, endowed with real causal powers – or, to put it differently, it presupposes that the position known as agent causation,
construed in substance causal terms, is true.3 If one were to assume instead either that agents’
seemingly intentional actions are, in reality, caused by something other than themselves (as per
both hard determinism and event-causal compatibilism), or that intentional actions escape or bypass causation in some way (as per event-causal libertarianism), the concept of alienation would
make no sense – not unlike the way in which the concept of an autoimmune disorder would be
unintelligible if there were no such thing as an immune system.
Let me begin with Marx’s approach in Capital to concrete laboring in capitalism, then move
to the concept of alienation. Marx treats even commodified labor-power as causally productive,
metaphysically. The labor-power that is sold for a wage is expended, by workers, in the transformation – by those same workers – of raw materials into new commodities, e.g., cotton into
cloth and then into shirts (Marx, 1992). At the level of metaphysics, the model is recognizably
Aristotelian: workers bring shirts into being by making them. Workers are the active, substance
cause of shirts, what Aristotle called the efficient cause. While it is the case that at time t cotton
cloth exists and at time t’ shirts exist, it is also the case that, no matter how tiny the time slices,
the order of succession (here, there and/or everywhere) is not what is taken to be the causal
factor. Doing by an agent or agents is: specifically, the form-giving of techné. That is why value,
as Marx conceived it, is a quantity of averaged generic efficaciousness, and not a quantity of
counterfactual dependence. Since causation is assumed to be productive – and since it is workers
who do the productive causing – one condition of capital accumulation is that workers have to
take in enough calories to be physiologically capable of playing the role that they do in the social
metabolic process of valorization, to use Marx’s metaphor. The concept of the wage is tied to this
requirement: the wage is the quantity of value, expressed as money (qua universal equivalent),
that is contained in the commodities that are necessary to ensure that workers will be able to
engage in literally energetic productive causation for a given period of time.
Moreover, it is not that Marx just happens to fall back upon Aristotle when it comes to the
physical process of creating artifacts. The idea that causation is productive, and that workers
are the efficient substance causes of the commodities that they make, is integral to his account
of capitalism as a social formation. Had Marx not assumed a productive, substance causal account of causation, then labor-power could not be said to be the source of surplus value – or of
any value, for that matter. Apart from the denial of agent causation undercutting the very idea
that it is the wage-relation that is the mechanism of exploitation in capitalist class societies, in a
more technical register we can see that if labor-power were not taken by Marx to be the source
of value, then the activity of workers would not figure into the valorization process as variable
capital, as opposed to fixed or constant capital. Etcetera. The Aristotelian metaphysics cannot be
jettisoned. But perhaps – or so one might think – we can allow that workers are the substance
causes of commodities, yet still deny agent causal free will. Marx stipulates that workers sell
their labor power willingly. While voluntary action is not the same thing as self-caused action, it
is unlikely that what Marx really meant when he insisted that it is a condition of wage-labor that
workers, unlike slaves, are free is that workers’ actions are literally caused, as per Harry Frankfurt’s fiendish Mr. Black, by capitalists. Still, maybe workers cause shirts to exist, but that very

substance causal activity is itself deterministically caused by something that is neither workers
nor capitalists. Or, maybe the act of entering into the wage-relation (albeit not the laboring itself)
is unfettered by causation altogether, in keeping with event-causal libertarianism. Nothing in
how Marx talks about the selling and expenditure of labor power invites either of these readings,
and I do not think that either is tenable, but if we isolate the issues of concrete laboring and/or
the selling of labor power, and consider them from the account as a whole, we might be able to
imagine, if only for a moment, that it is possible to have the concept of commodified labor-power
without implicitly assuming agent causal free will. As in the case of the atomist who replaces
claims about corporate entities with claims about individuals, here too one would no longer be
advancing Marx’s own metaphysics. However, assuming an event-causal position that is consistent with a belief in the reality of activity (no small presumption), one may at least entertain the
thought that there is nothing specific to the concept of workers selling or expending commodified labor power that is necessarily at odds with hard determinism, event-casual compatibilism
or event-causal libertarianism.
It is otherwise with the concept of alienation. Marx’s approach to alienation unambiguously
implies a commitment not just to productive substance causation, but to agent causal free will.
How so? Let us look more closely at the definition of alienation that I set out above. At the level of
false belief, alienation is a matter of not appreciating that value is our own total efficacy, averaged
and measured in units of time. It is efficacy that has been rendered generic, but value is us – or
at least those of us who are workers – and it is an error to imagine that it is not. This sense of
alienation plays a role in what Marx calls in the opening chapter of Capital ‘the fetishism of the
commodity and its secret (Marx, 1992).’ It appears to us (Marx writes there) that commodities are
exchanged for other commodities in ratios that are based on the utility of their natural properties,
as though exchange-value were a function of a demand for those properties, rather than being
a quantity of abstract socially necessary labor power. But the appearance is a mystification. In
reality, “the commodity-form, and the value-relation of the products of labor within which it appears … is nothing but the definite social relation between men themselves which assumes here,
for them, the fantastic form of a relation between things (p. 165).” Belief in the existence of metaphysically free agents as per an agent causal account is implied even by this limited, epistemic
sense of alienation. If determinism is assumed instead, then the agential causal powers that,
according to Marx, we have erroneously projected onto entities other than ourselves will not have
been ours in the first place. We might have been deterministically caused to imagine that they are
ours (all the while being deterministically caused to experience them as not-ours), but the reality
will be that we have no such powers. If event-causal libertarianism is assumed, meanwhile, then
our own actions will have to be sufficiently a-causal to not have been deterministically caused.
(And if there is any causing going on, it cannot be that it is done by us, else the view would collapse into a defense of agent causation.) Neither of these ontologies permits agents who have
misattributed their own real causal powers to entities other than themselves.
At a deeper level, alienation is not simply a mistaken belief about whose causal powers are
whose. We have in fact turned our powers against ourselves, Marx says, imposing upon ourselves
a logic of accumulation that prevents us from meeting even the most basic of human needs when
doing so runs counter to that logic. This non-epistemic sense of alienation also invokes, albeit
negatively, the idea that human beings are capable of self-determination not just socio-politically, but metaphysically. Were we not presumed to have the ability to initiate or decline to initiate
actions, we could not be thought to have turned that same agential capacity against ourselves
(just as we could not be thought to have misattributed it). It is hard to see how one who denies
the reality of agent causation – be she a hard determinist, an event-causal compatibilist or an

event-causal libertarian – could even describe the condition in which Marx claims that we find
ourselves.
Perhaps the determinists could say that alienation is when we have been deterministically
caused to imagine that things such as markets have real causal powers, and also deterministically
caused to defer to them, when in reality it is not the markets but something altogether different
that has deterministically caused our behavior. The compatibilist might want to add that for it to
be alienation we also have to have been deterministically caused to have the experience of not
wanting to do what we are deterministically caused to do (and do nevertheless do). But neither
of these claims approximates the idea that we have turned our own real agential powers against
ourselves. And it will be no better for the event-causal libertarian, who will have to say something
such as that alienation is when we experience what are in reality uncaused acts as being deterministically caused, presumably by prior events and so-called laws of nature.
At the deepest level, finally, the concept of alienation refers to the fact that value is not simply
a quantity of generic human efficacy, but rather a quantity of generic distorted or dehumanized
human efficacy. Value is in this sense inherently alienated. Its substance is labor power, as Marx
puts it, but the laboring that occurs in the process of valorization is no longer an expression of our
‘species being,’ viz., our distinctive ability to shape both the natural and the social world in accordance with the laws of beauty, as he put it in 1844, echoing Kant (Marx, 1964). Like the model
of concrete laboring, the concept of ‘species being,’ is recognizably Aristotelian. Human beings
are conceived as having essential properties – above all a capacity to self-consciously transform
not just nature but ourselves, i.e., for rational collective self-determination – the actualization
and expression of which is taken to be the ultimate good for creatures of our kind. This deepest
sense of the concept of alienation not only presupposes that agents have the capacity to initiate
causal processes of our own accord (i.e., that we are metaphysically free to do so), it also communicates both the depth and nature of the affront, from this perspective, of any version of the
idea (hard determinist, event causal compatibilist or event causal libertarian) that we do not
have such a capacity. There is simply no way to square Marx’s notion of the species-being from
which we are alienated with the idea that, as a matter of metaphysics, our every action is deterministically caused to occur by something other than ourselves (or otherwise that our actions are
necessarily non-causal). This is why the possibility that I held out above, viz., of taking up Marx’s
concept of commodified labor power but claiming that productive causal activity in the context
of capitalism is itself deterministically caused (or itself has no cause), is not viable. Laboring
under the value-form is not a proper expression of our essential human capacities, a condition
of energeia. But if estranged labor, as conceptualized by Marx, was laboring that is enslaved by
the metaphysics of causation itself (or if it were uncaused), then non-estranged, non-alienated
labor would also be metaphysically impossible. As it happens, Marx claims just the opposite of
that: non-estranged, non-alienated labor is not only metaphysically possible, according to Marx;
it is historically possible. Determinism might be true, of course. And if it is not, which I think is
more likely, it might well be thought to represent perfectly, at the level of thought, the material
reality of alienation. That might be its encoded sociological truth-content, as Adorno would put
it. But whichever way it is, you lose every sense of what Marx himself meant by alienation if you
do not assume agent causation, parsed as a species of productive substance causation, at the level
of metaphysics.
As before, with respect to the posited emergent phenomena that figure in works of Aristotle,
Rousseau and Marx, the point is not that Marx is correct in any of what he has to say about the
posited phenomenon of alienation. The claim is only that one cannot say what Marx himself says
if one denies the metaphysics upon which what he says is predicated.

4 | CONCLUSION
The moral of this story is not that everyone ought to stop what they are doing and become specialists in metaphysics. Nor – to be very, very clear – is it that social and political programs may
be read off of metaphysical commitments. On the contrary: I am in favor of all of us continuing
to work at the levels of abstraction that interest or concern us most, and by no means do I think
that a belief in the existence of emergent phenomena, for instance, tells us anything about which
emergent phenomena, if any, are the good ones, or that a belief in the reality of productive agent
causation implies anything about what we ought to do with ourselves. But the myth of metaphysical neutrality is false. Beliefs about which things exist, and what they are like, are built into
our concepts, our categories and our theories about the social world. I have focused upon hypothetical misalignments: professed ontological individualists who want to refer in their accounts
to emergent phenomena; event-casual determinists and libertarians who want to talk as though
there are agents endowed with the capacity to cause their own actions, and also to falsely project
and/or displace that capacity onto other entities. In practice, things are more likely to line up.
In the present environment, what is most likely is that one will have been trained to do social
science and/or social or political philosophy in a manner that presupposes the dominant metaphysics, key planks of which are ontological individualism and passivist, event-causal accounts
of agency. With respect to the former, what this means is that one is apt to either (a) restrict one’s
attention to the behavior and beliefs of individual persons (on the tacit assumption that emergent
phenomena do not exist); or (b) less coherently and coming closer to my examples, restrict one’s
attention to the behavior and beliefs of individual persons, yet believe oneself to be studying, say,
the institution of the state rather than state employees – or structural racism rather than individual prejudice. The situation is the similar with respect to the dominant event-causal metaphysics
of agency. Those who assume either that our actions are deterministically caused by prior events
or that they are not caused at all (let alone who think either of these things in conjunction with
thinking that causation is not productive) will simply be less likely to ask probing questions
about why and how our causal powers have been thwarted.
Plus, it is not just that the myth of metaphysical neutrality is false. Irrespective of the nature
of the metaphysics that it allows one to tacitly import into one’s work, subscribing to the myth
of metaphysical neutrality makes it less rather than more likely that one’s work will be critical.
This since it is in the nature of the case that it will not occur to one who maintains such a conceit
that their thinking may be circumscribed by an unwitting acceptance of one or another set of
implicit metaphysical commitments. For this reason, the myth is politically pernicious, and not
just epistemically pernicious. Specifically, by providing a meta-theoretical shield for claims about
the world that are tacitly shaped by metaphysical commitments which are both (a) functional for
maintaining existing relations of power, and (b) contestable, the myth of metaphysical neutrality
is not just false, but ideological. In “The Latest Attack on Metaphysics,” Max Horkheimer made
the point that the epistemological, methodological and ontological stance associated with 20th
century logical positivism was a “progressive movement” in the 17th century (Horkheimer, 1975).
But, he added, “it would be evidence of a most naïve interpretation of the historical situation to …
insist … that the force of the antithesis has not been shifted (p. 186).” Arguably, there was a time
when insisting that one could remain metaphysically uncommitted whilst making claims about
middle-sized dry goods was also progressive, even if it was never true that one could. But that
time, too, has passed. We are not all obliged to drop what we are doing and become metaphysicians, but we do all need to see through the myth of metaphysical neutrality. Paradoxically, the
claim that our social and political theories have no metaphysics built into them could only be

true if social phenomena were ultimately no way at all, or otherwise were all ways at once, such
that claims about them really could be consistent with any given underlying ontology. That idea
is itself a bit of fairly outlandish metaphysics, and there is no particular reason to think that it is
true.
DATA AVAILABILITY STATEMENT
This article contains no reference to data sets.
ORCID
Ruth Porter Groff https://orcid.org/0000-0002-3053-4068
ENDNOTES
1 An anonymous reviewer wonders if it might not “be part of a colonial erasure of competing ontologies” to expect
the content of a theory to be consistent with its underlying assumptions. I don’t think so. On the one hand, if the
implicit claim is that it is only Europeans who have the capacity to formulate coherent accounts of the world,
then the idea itself seems to me to inadvertently smack of colonialism, in addition to being false. On the other
hand, if the idea is that expecting theories to be consistent with their underlying assumptions is tantamount to
saying that there is no place for competing alternative (coherent) theories, the concern doesn’t follow. But even
if it did, I make no claim in the paper that a given ontology is correct – only that if one accepts certain beliefs,
one cannot consistently hold others. Finally, as I say in the conclusion, my focus on inconsistency is in large part
heuristic. In practice, the effect of the myth of metaphysical neutrality is primarily to generate theories that are
unselfconsciously consistent with unacknowledged, generally orthodox, assumptions. If anything, challenging
the myth should lead to more rather than to less room for substantive metaphysical discussion.
2 I use the word ‘whole’ to refer, paradigmatically, to any entity that has parts, in relation to which it is a question
whether or not the entity is something greater than or other than the sum of those parts. The standard example
in debates about emergence in the context of social theory is of a battalion and the soldiers that compose it. As
I say in the text below, some philosophers are of the view that wholes that meet the criterion of being emergent
(thereby) lose their parts. I continue to use the word ‘whole’ in talking about that view. However, I also make
use of the distinction between atomism and ontological individualism in order to specify more precisely that
one who holds such a view, while not being an ontological individualist, is nonetheless an atomist. In any case,
‘whole’ is not meant to imply ‘totality’ or ‘whole society,’ as my examples in the text make clear.
3 For an exhaustive discussion of what a powers-based approach to agent causation involves, see Groff (2019).
REFERENCES
Aristotle. (1998). In R. F. Stalley (Ed.,), E. Barker (Trans). Politics. Oxford University Press.
Bhaskar, R. (1978). On the possibility of social scientific knowledge and the limits of naturalism. Journal for the
Theory of Social Behaviour, 8(1), 1–28. In J. Mepham & D. Hillel-Ruben (Eds.), Issues in Marxist philosophy
(Vol III). Basil Blackwell. Reprinted in (1979).
Bhaskar, R. (1979, 1998). The possibility of naturalism. Routledge.
Durkheim, E. (1964). The rules of sociological method. The Free Press.
Groff, R. (2012). Ontology revisited: Metaphysics in social and political philosophy. Routledge.
Groff, R. (2019). Sublating the free will problematic: Powers, agency and causal determination. Synthese, 196(1),
179–200.
Horkheimer, M. (1975). The latest attack on metaphysics. In Critical theory: Selected essays. Continuum.
Marx, K. (1964). The economic and philosophical manuscripts of 1844. M. Milligan (Trans). International Publishers.
Marx, K. (1992). Capital, Volume I: A critique of political economy. B. Fowkes (Trans). Penguin.
Mill, J. S. (2006). On the logic of the moral sciences. A system of logic, raciocinative and inductive, Book IV. Liberty
Fund.

O’Connor, T., & Jacobs, J. (2003). Emergent individuals. The Philosophical Quarterly, 53(213), 540–555.
Rousseau, J. (2003). The social contract and the discourses. G. D. H. Cole (Trans). Everyman’s Library.
How to cite this article: Groff, R. P. (2021). The devil is in the categories: Metaphysics
and social and political thought. Journal for the Theory of Social Behaviour, 51(4),
675–688. https://doi.org/10.1111/jtsb.12320

Copyright of Journal for the Theory of Social Behaviour is the property of Wiley-Blackwell
and its content may not be copied or emailed to multiple sites or posted to a listserv without
the copyright holder’s express written permission. However, users may print, download, or
email articles for individual use.