1. Definitional Kind-Essentialism

Many have contributed to the systematic development of post-Kripkean essentialism (Plantinga [Reference Plantinga1974]; Forbes [Reference Forbes1985]; Wiggins [Reference Wiggins2001]; Almog [Reference Almog1991, Reference Almog1996, Reference Almog2003, Reference Almog2010]; and Hale [Reference Hale2002], among others), but to characterize definitional or categorical essentialism, I make use of Fine's work, which carries this view to its most extreme conclusions. In the vastly influential ‘Essence and Modality’, Fine (Reference Fine1994a) argues for the precedence and irreducibility of essence to necessity. This central tenet of Fine's essentialism has become common knowledge among contemporary metaphysicians. Let me briefly summarize some key points of Fine's view in relation to Kripke's text. Kripke, as is well known, argues for the necessity of identity. This, to start with, is simply the thesis that the logical relation of identity, that is, strict numerical identity, holds of necessity, that is, for any x and y, if x = y then necessarily x = y (Reference Kripke and Munitz1971: 136). Yet, under the necessity-of-identity banner, Kripke also argues for the necessity of theoretical identifications, like that necessarily water is H₂O and gold the element with atomic number 79 (Reference Kripke1980: 158–61). Yet, if water is deemed necessarily identical with H₂O, it must be because we reflect on what water is, we classify it as a chemical substance, and individuate chemical substances based on their chemical analysis. That chemical substances are so individuated is surely not part and parcel of the logical notion of identity. Similarly, gold is the yellow (suppose!) metal with atomic number 79, and let us grant that everything is necessarily what it is and not another thing. Still, the conclusion that gold is not necessarily yellow while it necessarily has atomic number 79 follows only if we grant the further assumption that chemical elements are individuated by their atomic number, not by their color: the element with atomic number 79 is what gold is. In Fine's work the unstated assumption behind Kripke's thesis of the necessity of theoretical identifications emerges explicitly: ‘I shall use the terms “essence” and “identity” (and sometimes “nature” as well) to convey the same underlying idea’ (Reference Fine and Sinnott-Armstrong1994b: 69, fn. 2). Essentialist necessities depend on identity understood as the nature or essence of things, not just on logical or numerical identity.

Still moving beyond Kripke, Fine unifies the metaphysical and the conceptual necessities under the essentialist heading. According to Fine, the metaphysical necessities depend on the nature, essence, or identity of all objects; the conceptual necessities depend on the nature, essence, or identity of all concepts. Moreover, the logical necessities are that subclass of the conceptual necessities ‘which are true in virtue of the nature of all logical concepts’, and similarly for other disciplines (Reference Fine1994a: 9–10). Thus, logical necessities too are essence dependent. No such overarching unification was explicitly envisioned by Kripke.

To better characterize essence, Fine (Reference Fine1994a: 10–16) revives the Aristotelian notion of real definition and contests the prejudice against real definitions as such, that is, against definitions of objects in particular as opposed to a general Quinean hostility against all definitions. Quine rejected the whole bulk of notions surrounding nominal definitions too: truth by convention (Reference Quine and Lee1936), analyticity, synonymy, and meanings ([Reference Quine1953] Reference Quine1961). But many are willing to endorse nominal definitions, yet balk at the idea of defining objects. Fine's argument against this double standard is based on the thesis that ultimately nominal and real definitions are of one and the same sort: the ‘nominal’ versus ‘real’ specification is a mere extensional classification by subject matter. These labels specify which entities, nomina or res, are defined. They do not indicate different kinds of definitions. This is reminiscent of Quine's (Reference Quine, Hahn and Schilpp1986b) classification of truths by subject matter alone, a point Fine seems to be applying to definitions and the correlated subspecies of essentialist necessities.

Fine's take on nominal definitions is reminiscent of another characteristic Quinean point: ‘Things had essences for Aristotle, but only linguistic forms have meanings. Meaning is what essence becomes when it is divorced from the object of reference and wedded to the word’ (Quine [Reference Quine1953] Reference Quine1961: 155). Thus, claims Quine, the supporters of analytic necessity succeed at best in restricting their essentialism to linguistic or conceptual entities: ‘For the appeal to analyticity can pretend to distinguish essential and accidental traits of an object only relative to how the object is specified, not absolutely. Yet the champion of quantified modal logic must settle for essentialism’ (Quine Reference Quine1951: 22). Evidently, Fine agrees with Quine on the essentialism hidden in nominal definitions. Meanings stand to words as essences stand to objects because meanings are essential to the individuation of words: a word is defined by its meaning. Fine (Reference Fine1994a: 13) also claims that meanings and concepts have essences too, as revealed by the fact that in specifying the meaning of a word we must characterize it too by its own essential features. Quine and Fine thus agree on the essentialist character of nominal definitions. But whereas Quine argued by modus tollens, from the rejection of essentialism to the rejection of nominal definitions and analyticity, Fine argues by modus ponens, from the acceptance of nominal definitions and analyticity to the acceptance of essentialism. And once essentialism is endorsed for a subclass of entities, what remains in place to stop its generalization to all entities? Fine (Reference Fine1994a: 13–14) is critical of such limitations: ‘On their view, it is only concepts and meanings which can be defined, and not objects. The difficulty with this position is to see what is so special about concepts.’

Is there really nothing special about concepts and no reason to recoil at the idea of defining objects but not concepts? Philosophers who, unlike Quine and Fine, discriminate nominal from real definitions seem to have some resources to do so. If definitions are understood as conventional stipulations, meanings and concepts can be regarded as the causal products of such definitions, making them a special sort of conventional artifacts. This view of concepts may be independently rejected, but in any case it is not obviously extendable to standard objects, which are not the products of conventional stipulations: we cannot by convention make an object (be what it is). If so, objects cannot be analyzed the way of concepts, that is, they cannot be deconstructed into their built-in, definitional components. Indeed, the basic distinction between a Carnapian analytic (as in Carnap Reference Carnap1956) and a Finean essentialist conception of necessity seems to reside exactly in this: in the analytic conception, definitions are conventional stipulations, making meanings up; in the essentialist conception, they are . . . what? To answer this question, let me return briefly to Naming and Necessity, and find there what I regard as the core seed of definitional essentialism, hidden in a casual remark of Kripke's.

To posit an invidious distinction between the essential and the accidental features of an object, we need a discriminating criterion. Arguing for the necessity of origin for Elizabeth, Kripke writes, ‘One can imagine, given the woman, that various things in her life could have changed’ (Reference Kripke1980: 113). This brief remark, I take it, encapsulates the core idea of all forms of essentialism: the essential features of an object are those that give the object and thus are prior to and independent from its life vicissitudes. In fact, the features Kripke necessitates are not really ways of being of their bearers. On the contrary, they are the preconditions for being given in the first place, viz., properties without which there would be no specific object to start with. Once Elizabeth is given, there are ways in which she is, and, in turn, there are alternatives to these ways, ways she might have been. But for this span of possibilities to be in place, Elizabeth must first be given. This seems to be the general underlying characteristic of all Kripkean essentialist necessities.

That is a general enough criterion, but in the specifics what is it to give Elizabeth? I believe we find in Naming and Necessity the outlines of two alternative answers to this question. The first answer well aligns with Fine's revival of Aristotelian essentialism and is encapsulated in Kripke's remarks on the distinction between (i) those questions that pertain to the identity and persistence of an object through time: ‘What properties must an object retain if it is not to cease to exist’ (Reference Kripke1980: 114, fn. 57), something he is not concerned with; and (ii) his topic of interest, the ‘non-temporal’ (his word) question that concerns necessity: ‘What (timeless) properties could the object not have failed to have, and what properties could it have lacked while still (timelessly) existing?’ (Reference Kripke1980: 114, fn. 57). The key word here is ‘timeless’. I suggest that this timeless notion of giving an object, even a material and history-bound object like Elizabeth or a table, is the central tenet of a definitional understanding of essence. To define an object is to give it in a timeless manner.

The timeless definition of a material object stands in sharp contrast to its material production. The definition of Elizabeth is not the time-bound causal making of that woman. A definition gives an object by spelling out (analyzing) its essence, nature, or identity. I propose the following as the central axiom of definitional kind essentialism (DKE):

The timelessness of essence as definition is emphasized in Fine's more recent work on essentialist necessity, which once again moves beyond Kripke in explicitly adding transcendence to timelessness. Truths of essence are taken to be both world and time independent. Thus, Fine (Reference Fine2005) distinguishes between (i) necessity as truth in all circumstances and (ii) necessity as truth independently from the circumstances. Essentialist necessity, claims Fine, is of the second sort as facts of essence are transcendental, that is, they are precircumstantial and as such independent of worldly circumstances:

Facts of essence are then deemed to be timeless exactly because transcendental. In attributing world-independence to truths of essence and the forthcoming necessities, Fine expands on Kripke's idea that the essential features of an object give the object rather than spell out ways in which the object is or even has to be: essential properties are not ways of being, not even inevitable, predetermined, necessary, sempiternal, or whatnot, ways of being. Yet, this development too goes well beyond Kripke's compressed suggestion. Fine's text makes it crystal clear that time-and-world-independent truths of essence cannot be regarded as actual truths—at least not in the standard worldly sense of ‘actual’. Essentialist necessities are transcendental truths. Actual truths are worldly truths par excellence: they are the truths about how things actually are. This is not just the relatively uncontroversial point that essential truths need not be actual because (some) essences need not be instantiated. Rather, essential truths cannot be actual because essential truths are transcendental, and the notion of actuality is an immanent, that is, worldly, notion. If transcendental truths of essence are not actual truths, obviously they are also not necessities in the sense of truths in all circumstances. Indeed, transcendental truths are not even possible in the ordinary circumstantial sense: ‘Truth does not imply possibility!’ quips Fine (Reference Fine2005: 327).

We have seen to what radical conclusions the ultimate rigorous development of the notion of definitional essence takes us. Essential truths provide the ontological scaffolding of the world and stand fundamentally apart from circumstantial truths, that is, from the immanent factual and counterfactual happenings of the world. Even though Fine does not state this explicitly, it is worth pointing out that sortal truths like ‘Socrates is a man’ and ‘Whales are mammals’ taken in the transcendental, categorical sense—the only sense in which they can be deemed essentialist necessities—have got to be as a priori as ‘Bachelors are unmarried’ and ‘Cicero is Cicero’. We have lost Kripke's famed a posteriori necessities! What we were supposed to know a posteriori is the actual truth that Socrates is a man, if such there be. But actual truths are not necessary in the intended essentialist sense. What is necessary is the transcendental truth of essence, an a priori counterpart of the Kripkean actual truth. In fact, strictly speaking, in Fine's assessment there is no tensed or actual truth of Socrates's being a man. Fine is explicit on the tenselessness of the predicate ‘is a man’, and the point must extend to modality: ‘is a man’ is a transcendental predicate. Indeed, being a man is not something that happens to Socrates; it is constitutive of his being given. As such it is a prerequisite for any actual vicissitude. Thus, insofar as actuality is a circumstantial notion, we must conclude that ‘Socrates is a man’ is not an actual truth as ‘objects enter the world with their identity predetermined . . . and there is nothing in how things are that can have any bearing on what they are’ (Reference Fine2005: 349).

Back in 1994 Fine had promulgated a unified essentialist interpretation of metaphysical and conceptual necessities. Conceptual necessities were exposed as nominal in name only but ultimately of the same essentialist brand as real necessities. Behind this metaphysical emphasis combined with the criticism of any partiality for nominal definitions lurked the promise of a realist account of all essentialist necessities. In the freshly post-Kripkean atmosphere of the time, perhaps led on by Kripke's much advertised a posteriori necessities, one might have been tempted to infuse a worldly character into Fine's essentialism. But once the transcendental character of definitions is clearly spelled out, metaphysical necessities reveal their true colors. The conventionalists may then perhaps be excused for favoring their own brand of necessity. Conventions can be regarded as human practices, generating a specific class of convention-dependent entities. The realm of essences, on the other hand, is meant to be neither conventionally stipulated nor worldly. Definitions and the forthcoming transcendental necessities are simply the (God-) given transcendental side of reality.

Which features of an object are transcendental, that is, constitutive of its essence? Fine argues for the transcendental nature of formal properties like self-identity and sortal properties, like man or set. It seems then that of the two main sorts of Kripkean necessities—kind membership on the one hand and origin with (derivatively) substantial makeup on the other (Kripke Reference Kripke1980: 115, fn. 57)— kind membership, that is, sortal classification, turns out to be fundamental to the definitional approach. Fine is explicit about this when defining the kind man: ‘The only plausible non-modal definition of “man” is to be an F (where this is the sort) differentiated in such and such a way’ (Reference Fine2005: 348). Formal properties like self-identity are also categorized as sortal. Fine even argues for a sortal sense of existence as the ultimate kind to which all things belong. In this sense, existence is a transcendental feature, written into the essence of all things (Reference Fine2005: 349–50). But, make no mistake, this is surely not what Quine might have dubbed the existential sense of existence. Nor does this sortal sense of existence guarantee the ordinary appropriate-to-its-own-kind, out-of-its-own-essence, existence of a thing, whether the transcendental existence that numbers and sets allegedly enjoy or the actual existence of Socrates. Of sortal existence, Fine says, ‘we might think of existence in this sense as the invariable concomitant of any object being what it is rather than of there being something that it is . . . to exist in this sense is simply to be an existent, i.e. the kind of thing that exists’ (Reference Fine2005: 353–54); thus, Fine concedes, such an existent need not be.

Whether Fine's sortal definitions deliver modal necessity, that is, a must, perhaps even in the form of genuine truth in all circumstances, and not just transcendental truth is a difficult and extremely interesting question that this paper cannot fully address. Of the modal import of the definition of man, Fine writes:

And:

These passages suggest that sortal properties alone are modal in the required sense, and predicates like ‘fleshy’ must be made sortal, as in the combined ‘fleshy-animal’, in order to deliver necessity. However, it is hard to see how even sortal attributes succeed in delivering necessity. Granted the understanding of sortal properties as individuative of what a thing is and precircumstantial, in which sense are such features necessary? One may raise the worry that no truth qua truth, not even truths of essence, can deliver the modal force of necessity (I so read Blackburn Reference Blackburn, MacDonald and Wright1986).

What about origin in the definitional account? Is it or is it not of the essence of Elizabeth to come from her parent-gametes? Are material objects individuated by their origin? Is origin perhaps their ultimate specific difference? Fine does not take a clear stand on mundane cases. He is however explicit on the real definitions of formal entities, like sets: ‘In specifying what a set is, we must state two things. First, we must state what general kind of thing it is—in this case, a set. Second, we must state how it is to be differentiated from other objects of the same sort—in this case, by its members’ (Reference Fine2005: 348).

A set can be completely individuated by its essence because set-membership is itself a formal relation as transcendental as the kind set; consequently, the transcendental definition of a set can completely specify it. For Fine, this holds for impure sets too. The essence of singleton Socrates can refer to Socrates as its member, without thereby leaving the transcendental realm. This is the case, despite Socrates himself not being transcendental, thanks to existence being written into his essence. Impure sets exist eternally and transcendentally and can still depend on the (transcendental and eternal) sortal existence of their non-eternal and non-transcendental members. The transcendental formal existence of {Socrates} and its dependence on Socrates are thus made compatible (Fine Reference Fine2005: 352–53).

It seems no accident that Fine's definitional essentialism remains silent on Socrates's mundane origination from a particular sperm and egg. Is the property of being so produced part and parcel of Socrates's nature? To answer this question in the affirmative one has to insert worldly, causal relations into essence. By all means this is not an impossible move, but one that goes against the transcendental and formal character of real definitions. Fine retains the formality of the definition of the singleton {Socrates} thanks to sortal existence being written into the essence of Socrates. We might similarly insert into the definition of Socrates a formal counterpart of his material process of origination. But to do so, more is needed than just the sortal existence written into the essences of the original sperm and egg. The material relation of origination, not just its relata, would have to be sublimated into a transcendental, formal counterpart. Ultimately, from the transcendental perspective, origin appears to be a highly circumstantial way of being as it is neither a sortal property nor the key notion of giving. On this point, see also Almog (Reference Almog1991 and Reference Almog1996), who emphasizes the distinction between the necessity of kind membership and the necessity of origin. In these early papers, Almog argues against Kripke that though both are necessary, only kind membership, and not origin, is an essential or whatness-pertaining feature of things. Almog's early misgivings about origin seem dictated more by its relational and non-sortal character than by its material nature. Fine's silence on origin is reminiscent of the early Almog's qualms and of Fine's (Reference Fine1994a) own early remarks on essence as nonrelational.

Additionally, Fine distinguishes between the transcendental core of the essence of an object and its complete essence:

Material production, unlike set membership, is not a formal relation, and as such it cannot belong to the transcendental core of Socrates's essence. Yet, Socrates's full individuation seems to require (the mention of) this worldly relation—or of some alternative material source of individuation. That is, either Socrates's transcendental essence is nonindividuative (and it remains to be specified how it includes the promised blueprint of his full essence), or it must include a formal counterpart of origin or of some other material relation, for example, composition. No matter which option we pick, the real material relation of origination does not appear to belong to what Fine calls the real, but ought perhaps more properly have called the formal, definition of a thing. I suspect that Fine's idea of an extended notion of essence springs exactly from this difficulty: insofar as the full individuation of material objects calls for material relations, one must renounce either the pure formality or the completeness of their definitions.

The crucial point is that in the transcendental perspective the historical production of a thing can be necessitated only if it can be formally represented and inserted into its essence as the specific difference in its categorical definition. To paraphrase Fine (Reference Fine2005: 348) on sets, we can say of men: ‘In specifying what a man is, we must state two things. First, we must state what general kind of thing it is—in this case, a man. Second, we must state how it is to be differentiated from other objects of the same sort—in this case, by its origin.’

Let me briefly summarize our discussion of Fine's definitional essentialism. First, despite Fine's protests, there seems to be some space to judge real definitions as more problematic than nominal ones, at least if we are skeptical of a transcendental Platonic realm of essences, in contrast to a convention-based realm of meanings. And interestingly, for all that has been said, a fragmentation of reality into irreconcilable transcendental layers of clashing systems of definitions carving nature at incompatible joints has not been ruled out. Second, real definitions are definitions of res, but they are still formal ways of giving such res. As we have seen, in the definitional sense of giving, given objects need not exist though a categorical counterpart of existence can be integrated into essence as the ultimate, most general form of sortal categorization. Third, and this will come as no surprise, the connection between the transcendental and the worldly layers of reality begs for an explanation, especially the linkage between essences and their corresponding objects, but also that of transcendental truth with actual truth, on the one hand, and of essentialist modality with circumstantial modality, on the other. Finally, the question of whether we can derive modal necessity (a must) from essence remains open. On this last point, one might wonder whether essentialist necessity is nothing but transcendental truth.

2. Historical Origin-Essentialism

Kripke's argumentative efforts—both his informal arguments, as in the main text of Naming and Necessity, and quasi-formal proofs, as in the much-discussed footnote 56—focus on the necessity of origin, rather than on kind membership. Informally, Kripke contemplates Elizabeth and says:

Of a wooden table we are told:

Notice that Kripke is not inviting us to imagine a different origin for the already given Elizabeth or table as if he were to ask, while pointing at Elizabeth and at a separate (picture of a) couple of gametes, whether something prevents us from conjecturing that those are her parent gametes. To this the answer seems a resounding no: nothing so prevents us. But his request is rather to imagine ‘her being born of different parents’, and we are told that ‘though we can imagine making a table out of another block’, this will not result in imagining this very table. The request is thus not so much to imagine that the relation of origination holds between two statically and separately (re)presented or imagined objects; we are instead invited to imagine the dynamic process of production leading from one object to the other, as Kripke's use of the progressive tense makes clear.

These passages seem key to addressing the difficult question of whether we can imagine or conceive the (alleged) impossible, for example, water's chemical composition being H₂O₂ or Elizabeth being an offspring of the Trumans’. To answer this question, we need to set aside propositional imagining and conceiving, that is, imagining (de dicto) that x is F, or even (de re) of x that it is F. Is it possible to imagine, wrongly as it turns out, that water is H₂O₂ or that Elizabeth is a biological daughter of the Trumans’? Similarly, is it possible to imagine of water that it is H₂O₂ and of Elizabeth that she is a daughter of the Trumans’? The answer to these questions seems positive if we can secure, for instance by causal means, that it is of water and of Elizabeth that we are imagining the impossible. What on earth prevents me from looking at Kripke's wooden table and think of it that it is made of ice? This, however, must be distinguished from the nonpropositional, objectual imagining and conceiving with which the above passages seem concerned. We are asked not so much to imagine that this table is made of ice, but rather to imagine this table as being made of ice. Kripke's question is whether we can succeed in imagining water itself or Elizabeth herself when we imagine, for example, H₂O₂ or the offspring of (some of) the Trumans’ gametes. To these questions, claims Kripke, the answer is no, given that to imagine H₂O₂ (making up a substance) just is to imagine hydrogen peroxide, not water, and to imagine the transformation of the Trumans’ gametes into a baby just is to imagine the offspring of those gametes, not Elizabeth, namely, the offspring of her own parent gametes. I surely fail to imagine water or Elizabeth if I imagine a distinct chemical substance or offspring, because, Kripke must assume, H₂O, not H₂O₂, is what water is and the offspring of her original gametes, not of the Trumans’, is what Elizabeth is.

Be that as it may, the focus in these passages is not on a timeless giving of an object but, on the contrary, on its temporal process of production. This focus on time returns in footnote 57 (the same footnote stating that essence is timeless!):

This last remark makes it clear that the focus on time is not a simple byproduct of the specific time-bound relation of material production under discussion. Rather, time and historical development are deemed to be part and parcel of ‘a general principle about essence’. This is clearly a temporal alternative to timeless essence. If what is possible for an object is what an already made object can sustain (a vicissitude of that object), an alternative making of the object does not count as a possibility for it. This suggests the essentialist necessitation of those features of an object that pertained to its temporal origination, rather than to its categorical individuation by definition.

From Kripke's remarks we can lift an alternative historical and origin-centered rather than transcendental and kind-centered form of essentialism. Elizabeth and the table are historical res. Such objects are certainly neither defined into existence in a God-like manner nor formally defined in a set-like way. To give Elizabeth, as opposed to a transcendental essence thereof, is to make Elizabeth herself and not to provide the essentialist blueprint of the transcendental requirements any object has to meet in order to be Elizabeth. Without the filter of transcendental essences, objects are actually given by the worldly and temporal processes of production that literally speaking make them. Thus, the giving of Elizabeth is nothing but the temporal causal process that brought her into existence. In this historical perspective, an object is essentially (given as) the product of its process of production. This immanent interpretation of essence is no simple modal view as it can still distinguish the essential prerequisites for generating Elizabeth, from (i) the features that Elizabeth cannot lose while still continuing to exist and, more generally, (ii) all the accidental circumstantial necessities involving Elizabeth, that is, what must hold in all circumstances in which she exists. Origin, that is, the prerequisites for existence, appears to give us a natural demarcation point between the essential and the accidental truths pertaining to an object.

This, of course, does not resolve specific origin essentialist theses insofar as the question of what exactly counts as the origination process of a specific object is not settled. For example, it is an additional question, independent of the general origin essentialist perspective, whether the Elizabeth-making process started with a pregnancy or with the Big Bang. Surely, we may endorse definitional essentialism and yet disagree on which features of an object are constitutive of its transcendental definition. Similarly, we may endorse origin-centered essentialism but disagree on which past events count as part and parcel of an object's process of production or on which other things (gametes or garters) are constitutive of its making.

In historical essentialism, the temporal giving of an object (its origin) plays the role that the formal giving of an object (its definition) plays in transcendental essentialism. Hence, the essentiality of origin is not a particular essentialist thesis that needs to be proved. The essentiality of origin is, so to speak, the fundamental axiom of the historical conception, which takes objects to be individuated by their processes of production. The original production of an object is the historical counterpart of definition. On Kripke's behalf, I propose the following core axiom of historical origin essentialism (HOE):

Interestingly, in the same footnote where history takes center stage, Kripke endorses the possibility of vague essentialist propositions:

This vagueness is perfectly at home in a historical perspective where what it took to make Elizabeth was not just a pair of gametes. The gametes were not magically transformed into a baby. To make a baby out of a pair of gametes, we also need an approximately nine-month-long process of gestation that requires food, water, and energy in the right womb-like environment. Similarly, even in the simpler case of a water molecule, it takes more than just the right kinds of atoms to form it. The atoms need to bond in the proper way. Chemical bonding, like fetal development, is a historical process. But then some vagueness on what exactly the process of production consists in is to be expected. No such vagueness is naturally forthcoming in the definitional model though, of course, like origin itself, it can be incorporated into it.

The time-centered historical essentialism of the main body of Kripke's footnote 57 owes much to Arthur Prior's (Reference Prior1960, Reference Prior1962) work. This is evident in the background assumption of possibilities as open future alternatives, that is, as time branches. Not only is historical essentialism embedded within a Priorean framework for possibilities, the specific idea that origin precedes classification by kind membership is also anticipated by Prior who so objected to the suggestion that it pertains to the nature of a specific kind of entities—acts of volition or hydrogen atoms—to come into existence uncaused (Reference Prior1962: 57): ‘It just won't do . . . to say that it is the peculiar and special nature of volitions to start into being without a cause, for nothing has any nature until it is there, so that whatever a thing's nature may explain or permit, it cannot explain or permit the thing's starting to be’. In the historical perspective, gone are Kripke's timeless and Fine's transcendental essences or natures. The whole idea that essence precedes existence is reversed—at least in the temporal order, but it is not clear that for Prior there is a higher order than that. The nature of an object is now seen as posterior to the object itself. What remains in place to precede and give the object is simply its origin. It seems then no accident that Prior (Reference Prior1960: 688) attempted his own proof of the necessity of origin, whose core idea is that at no time might an object have had a different origin: after the object exists it is too late; before the object existed nothing was yet possible for it.

The most articulate development of historical essentialism can be found in Almog's Reference Almog2003 and Reference Almog2010 papers. Although Almog (Reference Almog2010) with some good reasons rejects the term ‘essentialism’ as wedded to the definitional strand, he still endorses an origin-centered understanding of necessity: ‘Necessary conditions for x are byproducts of the process that sufficed for Nature to produce x’ (Reference Almog2010: 377). Almog's particular form of historical essentialism stresses the role of efficient causation and the interrelated nature of all things, and indeed a thing can perhaps be defined in isolation, but ordinarily it surely cannot be so generated. Two additional aspects of Almog's view are worth emphasizing. First, Almog regains the essentiality of kind membership by incorporating it into origin: ‘It is because their generation processes are different that they, the products, are of different kinds’ (Reference Almog2003: 205). If so, the essentialist necessity of ‘Socrates is a man’ does not depend on the kind man being a sortal classification. It depends instead on Socrates's material process of origination. The process that gave origin to Socrates is part and parcel of the way in which the kind man, understood now not as a sortal category but as the human species itself, propagates, and so Socrates came to be as a member of that species. We discussed how in the definitional perspective to be classified as essential, origin needs to be written into a definition. In the historical perspective it is kind-membership instead that must be regained as an aspect of the process of origination. Second, Almog argues for the generalization of the origin-centered paradigm beyond material objects. Mathematical objects, like sets and numbers, are reconceptualized as generated rather than defined. Here too, we see that the historical model runs counter to the definitional, whose central cases are pure abstract entities. In a similar spirit, we may even mirror within the historical paradigm Fine's classification of the conceptual necessities as a subclass of the essentialist necessities. If concepts, too, are part and parcel of (the cognitive corner of) the natural world, the door is open to claim that the concept bachelor is made when the concepts man and unmarried come to be psychologically associated in the concept-making way.

Let me briefly summarize our discussion of time-centered historical essentialism. First, a terminological point. The term ‘essentialism’ is not ideal insofar as essences, qua transcendental principles of individuation, are not relevant to this doctrine. Yet, it is still appropriate as this doctrine, too, endorses essentialism's most fundamental tenet, namely, the necessitation of the object-giving features. Both doctrines share the key idea that essentialist necessities are linked to the (precircumstantial) ontological structure of the world, not to the circumstantial possibilities for objects. An object needs to be given before circumstantial possibilities arise for it. Thus, in both views essentialist necessities capture the pre-modal ontological scaffolding of the world. Their disagreement is not so much on essentialist necessity per se, but rather on the nature of ontology. Definitional essentialism regards ontology as transcendentally fixed; historical essentialism takes the world to be, in and of itself, a world of (worldly-made) objects. Second, in this perspective the existence of an object precedes its categorical classification because an object must exist in order to have a nature. However, it is not to the nature—if such there be—of an object that essentialist necessities are now due. Rather, the material (preconditions for the) production of an object is what gets to be necessitated. Third, this form of essentialism posits neither transcendental essences nor transcendental truths. If all truths are immanent, essentialist necessities are just a subclass of actual worldly truths, and no gap is there to be filled between the essential and the actual truths or between objects and their essences. Finally, however, like in the definitional model and perhaps more urgently so, the question of whether we can derive a must from essential truths remains unanswered. In which sense are truths of origin necessary? Surely, there are no circumstances in which Caesar can proceed to have different parents because his process of production precedes his very existence, but wherein lies the modal force? In this perspective, it is the truths of origin for an object x that are deemed to be precircumstantial-for-x because x must be made before anything can happen to it. Precircumstantial truths are transcendental for Fine, and transcendental truths may well form an invidious subclass of metaphysically privileged truths. But in the historical perspective, there is no place for either a special class of entities (essences) or the special truths that pertain to them. If so, how can the process that sufficed for nature to produce x deliver necessary conditions for x? I turn now to this pressing question.

In the historical framework, origin is deemed necessary as an object is taken to be given by its origin. One crucial question remains open: what exactly counts as the origin of an object? What does the process of origination of an object consist in? And which of its aspects are, so to speak, rigid? Prior's proof is standardly faulted for proving too much. If it was at no time possible for Caesar to have different parents, then it is also the case that it was at no time possible for Caesar to be conceived in a different location. And for the same reasons, at no time was it possible for Caesar never to have existed. This line of objection focuses on the difficulty for the historical conception to necessitate exactly what we pretheoretically want to necessitate, for example, the parent-gametes as opposed to the location of conception. In line with these pretheoretical intuitions, Kripke (Reference Kripke1980: 115, fn. 57) claims that ‘the time in which the divergence from actual history occurs may be sometime before the object itself is actually created. For example, I might have been deformed if the fertilized egg from which I originated had been damaged in certain ways even though I presumably did not exist at that time.’ But such a retreat can hardly be justified in the historical conception. One way to justify why the making of Elizabeth requires her parent-gametes (though they too, like Kripke's fertilized egg, might have been tampered with) but not, let us say, gestation in her mother's own womb, consists in pointing out that this is what Elizabeth is: the product of those gametes. Unfortunately, this notion of ‘what Elizabeth is’ seems to backslide into a transcendental individuation of Elizabeth insofar as it privileges some aspects of the process of origination over others by defining Elizabeth as the outcome of those selected aspects, for what is whatness if not essence by another name? Without recourse to the nonhistorical, definitional notion of whatness, one can only conclude that Elizabeth is the product of her actual process of production, warts and all. This view seems stuck in a dilemma: either prove too much, Prior's way, or resort to considerations that are external to the sheer historical conception.

Moreover, we can grant the distinction between what we may call the pre-Elizabeth (precircumstantial for Elizabeth) truths concerning her origination and the post-Elizabeth circumstantial truths concerning her life vicissitudes. Yet, this is not a distinction between two different sorts of truths. And there is nothing in the pre-Elizabeth truths that promises necessity. The pre-Elizabeth truths are nothing but ordinary actual truths pertaining to the making of Elizabeth. Historical essentialism thus turns out to be an unstable position. One may sift the desired essential from the nonessential truths of origin by appealing to some independent principle of individuation, thus gaining, if not exactly necessity, at least a special sort of truth as well as the coveted extensional match with our pretheoretical intuitions, but at the cost of leaving the purely historical plane. Alternatively, in and of itself, historical essentialism cannot discriminate between the alleged essential and the nonessential aspects of origination. Also, it has no resource to make all truths of origination necessary, transcendental, or in any other way privileged—something somewhat unsurprising in a naturalistic conception. Relevant to these considerations are perhaps the intriguing remarks of Foucault who opposes a genuine genealogy to an essentialist quest for origins (Reference Foucault and Bouchard1977: 142): ‘Why does Nietzsche challenge the pursuit of the origin (Ursprung), at least on those occasions when he is truly a genealogist? First, because it is an attempt to capture the exact essence of things, their purest possibilities, and their carefully protected identities, because this search assumes the existence of immobile forms that precede the external world of accident and succession.’ To the genealogical conception we now turn.

3. Genealogy and the Origin of Essentialist Necessity

Foucault's remark quoted at the end of the preceding section contrasts an essentialist idealization of origin with a genuine genealogy. In a speculative vein, this final section briefly, all too briefly, explores Nietzsche's genealogical methodology in opposition to an essentialist methodology. To exemplify the two methodologies, I oppose some of Nietzsche's ([Reference Nietzsche1887] Reference Nietzsche1997) remarks on justice as fairness to Rawls's (Reference Rawls2001) influential account. Once the contrast is uncovered, I conclude the paper with a (very!) speculative suggestion on the genealogy of our concept of essentialist necessity itself.

Both Rawls and Nietzsche appear to attend to the phenomenon of justice as fairness. Yet, their philosophical projects could not be more radically opposed. What differentiates them is not only their final theories, but their entire methodologies. Rawls aims to justify a certain conception of justice by rationally deriving a system of fair principles from an ideal impartial setup where all morally destabilizing information, like that I am weak and you are strong, is hidden from our consideration. Nietzsche's aim instead seems to be (i) to describe the material conditions in which justice systems come to be established and (ii) to unmask the humble historical origins of our values in the very real and, if you want, accidental power struggles between very differently endowed creatures. These opposing methodologies are appropriate to two distinct subject matters. Rawls's methodology consists in ‘striv[ing] for a kind of moral geometry with all the rigor which this name connotes’ (Reference Rawls2001: 105). The starting point from which Rawls's conception of justice as fairness develops, is an ideal original position ‘in which the parties are equally represented as moral persons and the outcome is not conditioned by arbitrary contingencies or the relative balance of social forces’ (Reference Rawls2001: 104). The aim is to arrive by means of ‘strictly deductive’ (Reference Rawls2001: 104) arguments at the principles that make up a concept of justice that all the free, rational, noumenal selves represented in the initial position would subscribe to (see Reference Rawls2001: 225). The subject matter of Rawls's theory then is not so much a historical phenomenon, let us say, Western systems of justice in the twentieth century, but rather the fundamental principles that define the essence of the concept of justice.

On the other hand, Nietzsche's genealogical conception is based on the insight that ‘there is no more important proposition for every sort of history than . . . that the origin of the emergence of a thing and its ultimate usefulness, its practical application and incorporation into a system of ends, are toto caelo separate’ ([Reference Nietzsche1887] Reference Nietzsche1997: 51). Thus, if our purpose is to understand the historical emergence of the actual practice of justice, no rational justification (incorporation into a system of ends) needs to be provided. In fact, Nietzsche explores the ‘Origin of Justice’ and agrees with Thucydides that ‘justice (fairness) originates between parties of approximately equal power . . . where there is no clearly recognizable superiority of force and a contest would result in mutual injury . . . the characteristic of exchange is the original characteristic of justice’ ([Reference Nietzsche1887] Reference Nietzsche1997: 124). It seems then that to understand the historical and social phenomenon of justice, we must attend to exactly those natural contingencies that are the main impediment to Rawls's project of rational justification. Also, Rawls argues that ignorance of our real social position is required to reach the right conception of justice. Nietzsche instead speculates that it took our awareness of a real and therefore approximate balance of power actually to induce us to reciprocal fairness. (Sadly enough, in both accounts justice seems to be rooted in fear: be it the fear of picking a norm that may put us at a disadvantage or the fear of confronting a party we have no hope to overcome.)

In his Genealogy, Nietzsche not only uncovers the natural history of some human moral practices, like justice as fairness or punishment, he also uncovers the natural history of our moral concepts.Footnote ¹ As Clark points out, ‘Genealogy is simply a natural history. If there is something new in Nietzsche's use of genealogy, it is the suggestion that concepts are formed in the same way as other living things—and in particular that this is true of the concept of morality’ (Reference Clark2015: 31). Thus, our concept, ideal, or value of justice, which Rawls's theory aims to analyze, is for Nietzsche another historical entity over and above particular systems of justice, whose natural psychological development may be explored. In fact, the main body of the Genealogy focuses on the historical emergence of moral concepts.

According to Nietzsche's naturalistic perspective, a genealogical investigation is the proper methodology for understanding a historical phenomenon in and of itself, whether an object like Elizabeth or a social practice like fairness. A genealogy unearths the historical emergence of the phenomenon with no ulterior aim of uncovering its ultimate nature because ‘only something which has no history can be defined’ ([Reference Nietzsche1887] Reference Nietzsche1997: 53). This is a total reversal of the general definitional philosophical methodology shared by Fine and Rawls, important differences aside. But the genealogical conception must also not be confused with historical essentialism and its idealization of origin as ultimately revelatory of what an object is and its concomitant necessities.

Nietzsche reinterprets moral values too in a naturalistic perspective, offering a natural history of their psychological emergence. I am in no position to offer an analogous explanation of the psychological emergence of the concepts of essence and essentialist necessity. Let me only suggest that a natural psychological tendency to explain historical phenomena by attending to their genealogy may, so to speak, ‘degenerate’ into an inflated notion of origin as revelatory of what a thing really is and then lead to the further idealization of ‘whatness’ into a transcendental notion of essence. In this perspective, definitional essences are the idealized counterpart of historical processes of production. Of his moral investigations, Nietzsche says ([Reference Nietzsche1887] Reference Nietzsche1997: 4–5): ‘I was preoccupied with the problem of the origin of evil. . . . I learnt, in time, to separate theological from moral prejudice and I no longer searched for the origins of evil beyond the world. . . . Some training in history . . . soon transformed my problem into another: under what conditions did man invent the value judgements good and evil?’ This passage can be mirrored for essentialist necessity in place of evil. First, a metaphysician looks for the source of necessity beyond the world in transcendental essences. We then learn to look for the source of necessity in the world itself, descending from transcendental to historical essentialism. Finally, our problem is transformed into another: under what conditions did man invent the value judgments of essential versus accidental truths?

One point of this paper is to interpret Nietzsche's invitation ‘to kick to pieces the rotten armchairs’ ([Reference Nietzsche1887] Reference Nietzsche1997: 117) as an invitation to undertake a genealogical investigation of historical entities, descending, so to speak, from definitional essentialism through origin essentialism to a sheer genealogical investigation. I have focused on Kripke's cases, like people and water, to illustrate these three theories and suggested that we renounce essentialist necessities and replace them with plain truths of origin. As the title suggests, however, my ultimate concern has been to sketch a genealogical reinterpretation of the concept of essence itself, rather than a definition thereof. A genealogist, in contrast to a Finean metaphysician, will not attempt to define the notion of essence, but in the case of essence, too, they will investigate the historical emergence of this notion. Nietzsche's genealogical investigation of the practice of justice leaves unsatisfied any hope of finding something nobler than the balance of power at its core, and his investigation of moral concepts leads us back to their unflattering psychological roots. In a similar vein, a genealogical investigation of essentialist necessity leads us down from the Platonic or Kantian realms to the humble truths of origin and to the human creation of transcendental metaphysical notions.