What is traducianism?

One of Friedrich Nietzsche’s more plausible observations was that historically individuated ideas are hard to analyse; ‘it is only that which has no history, which can be defined’ (Nietzsche 1887/Reference Nietzsche1913, 93–94). Applied to the notion of traducianism, it is difficult to extract its essential claims from the broader theories of its proponents. I shall attempt to do this based on what the view’s paradigmatic ancient and medieval expressions have in common, and on how traducianism has been contrasted with its alternatives when ancient authors surveyed the options.

It is worth noting that though for most of its history, traducianism has been discussed as a chiefly theological doctrine, it had a half-millennium-long secular prehistory going back to the time of Aristotle, and counting among its defenders members of the Aristotelian, Stoic, and even Platonist schools (Kitzler Reference Kitzler, Hunter, van Geest and Peerbolte2024, 359–360; Falcon & Roreitner Reference Falcon and Roreitner2024; Roreitner Reference Roreitner2024). Even so, these discussions have exerted meagre influence and remained difficult to access; in the case of Aristotelians and Stoics, it isn’t even clear to what extent they can be understood in substance dualist terms. Therefore, I shall restrict myself to the analysis of Christian treatments, but without depending on their theological convictions for my own defence.

In his A Treatise on the Soul, Tertullian (c. 155–220) puts forth an account of the soul’s origin which is widely held to be the first Christian articulation of traducianism. Being some sort of a dualist, he argued that reproduction must involve ‘two kinds of seed – that of the body and that of the soul’, but he insists that these ‘are inseparable, and therefore contemporaneous and simultaneous in origin’. He then seems to suggest that these two ‘seeds’ are both originally parts of the father: ‘the soul-producing seed … arises at once from the out-drip of the soul, just as that fluid [which] is the body-producing seed … proceeds from the drainage of the flesh’ (Tertullian Reference Tertullian, Roberts, Donaldson and Coxe1885, A Treatise on the Soul, ch. 27).

Fourth-century theologian Gregory of Nyssa’s discussion cannot be abbreviated much without loss:

Note that Gregory’s roundabout phrases ‘that which is implanted for the conception of the body’ and ‘the thing so implanted’ refer to the male seed; it is thus clear that for him, the sperm contains everything, at least in potential form, that is necessary to produce both body and soul – ‘not by the importation of any other power from without, but by the power which resides in it’.

Jerome, writing in 410 CE, raises a list of possibilities on the origin of souls. Corresponding to traducianism, he writes the following: ‘or are souls really produced … by propagation, so that as the body is the offspring of body, the soul is the offspring of soul, and exists on conditions similar to those regulating the existence of the inferior animals’ (Jerome, Letter 126, 1.1).Footnote ² In 419 CE, Augustine insists on the disjunction between traducianism and creationism in the following words: ‘it is one of two things: if the soul is not derived by natural descent from the parent, it comes out of nothing’ (Augustine of Hippo Reference Schaff1887b, On the Soul and its Origin, I.24).

Traducianism boasts only one notable advocate in the Middle Ages. In a detailed treatise on original sin, working with a hylomorphic framework, Odo of Tournai affirms it in the following words:

These expressions of traducianism differ in multiple ways. First – and we shall do our best to set this aside – they likely differ in the details of what they take souls to be and what powers they attribute to them while in the womb. Second, only some of them (most clearly, Tertullian) endorse the notion that the offspring’s soul depends materially or at least constitutively on the father’s; that the former literally used to be a part of the latter. Third, they seem to take divergent positions on whether the soul is passed on through the material sperm (Gregory of Nyssa, apparently) or, in parallel but distinct fashion, through another, invisible ‘seed’ (likely, Tertullian).

These divergences lead me to reject aspects of some contemporary characterizations of traducianism. Contra the second point of divergence, Crisp (Reference Crisp2006, §§35–49) and Farris (Reference Farris2015, 381–385; Reference Farris2016, 45–47) suggest that traducians are committed to souls either splitting off (fission) or giving off parts (parturition) to explain how children’s souls come into existence. But as we’ve seen, this is not part of every paradigmatic characterization of traducianism, and so we cannot rule out accounts which don’t make such a claim. (At the same time, Leidenhag is right to note that otherwise ‘the traducianist faces a “lack-of-theory problem”’ (Reference Leidenhag2022, §2.1); later sections of this paper shall try to remedy that.) Contra the third point of divergence observed, Kitzler’s otherwise apt description unjustifiedly insists on the exclusive role of the material seed: ‘the soul, germinally contained in bodily sperm, is transmitted through sexual intercourse from the parents to their offspring’ (Reference Kitzler, Hunter, van Geest and Peerbolte2024, 359).

What the surveyed accounts do agree on, I submit, is largely captured by the following three theses:

A few comments are in order. First, in the context of the development of traducianism, ‘naturalness’ contrasts with supernatural intervention. Thus, an account is natural if it only appeals to natural entities and the laws of nature. Second, what it means for the child’s soul to depend on the parents is left vague, since the consensus of paradigmatic traducians in history leaves it vague. One thing it surely entails is that no natural account of human reproduction and ontogeny can explain the origin of the child’s soul unless it involves the parents. Third, Naturalness successfully contrasts traducianism with standard versions of creationism, while Parental dependence contrasts it with standard versions of emergent dualism – though it could be argued that this latter contrast is anachronistic in nature. Fourth, by proposing Parental dependence, I shift the focus from the male to both sexes. This is unapologetically revisionary; the male focus in ancient and medieval traducianism likely stemmed from an ignorance of much reproductive science. Fifth, Immediate ensoulment serves as a clarification or constraint on the other two conditions, ruling out ideas inconsistent with traducianism such as the pre-existence of the soul or delayed ensoulment resulting from a process initially triggered by the parents. One might argue that Immediate ensoulment is entailed by Naturalness or by Parental dependence – I shan’t pursue this matter.

One limitation of this discussion should be acknowledged: traducianism as put forth historically pertains only to sexually reproducing species. If some asexually reproducing animals have souls, they’d need a different account.Footnote ³ No doubt traducianism-inspired models could be offered that still afforded a role to the parent or, so to say, the original organism in the production of the offspring’s soul – but that is beyond the purview of this paper.

Traducianism for emergentists

Emergent dualism was introduced into recent discussion by William Hasker (Reference Hasker1999, ch. 7) and has also been endorsed – whether in those very words or in substance only – by Nida-Rümelin (Reference Nida-Rümelin, McLaughlin and Cohen2007), Zimmerman (Reference Zimmerman2010), Meixner (Reference Meixner, Nagasawa and Matheson2017), and Swinburne (Reference Swinburne2019, ch. 7). The core of the view is that substantial minds are brought about by, but not reducible to, sufficiently organized brains, nervous systems, or bodies; that minds remain dependent on their brains for their operations (and perhaps their existence) at least during their earthly careers; and that all of this is governed by laws of nature often called ‘emergent laws’. This dependence-without-reduction is held to be the distinguishing feature of ‘strong emergence’, accounting both for its advantages and for the suspicions that surround it (O’Connor Reference O’Connor and Zalta2021). In a recent chapter, Hasker summarizes the view as follows:

Consider now two versions of emergent dualism:

I shall now argue for two claims. First, that Parental-gametes emergent dualism (henceforth: PGED) is a version of traducianism. Second, that PGED is no less plausible than Own-body emergent dualism (henceforth: OBED).

To see that PGED is a version of traducianism, consider the three theses with which the view was characterized in the previous section. PGED meets Naturalness since, according to emergent dualism, the emergence of souls is held to be a part of nature; it is governed by lawlike regularities and requires no divine intervention. But two humans jointly having the power to produce a soul doesn’t seem any less natural than a single human body having such a power; after all, the power to produce a body is distributed, in some way, between creatures of opposite sexes. That PGED meets Parental dependence and Immediate ensoulment is clear, since it is the parents’ gametes that give rise to the new soul on the occasion of fertilization.

Let us now turn to the question of plausibility. On OBED, the body’s (or the brain’s) power to create a soul is considered strongly emergent: it cannot be reduced to, or understood as a mere combination of, the powers of the body’s or the brain’s constituents. But strong emergence has an importantly mysterious aspect to it. What distinguishes it from weak emergence is that what emerges isn’t entailed by what it emerges from – thus, we cannot predict a priori either the presence or the absence of the emergent phenomenon. However, this means we cannot protest against suggestions of strong emergence based on the inadequacy of the base. This is important because PGED suggests that the power to produce a soul is mediated by the gametes (the sperm and the egg) – perhaps less intuitively fitting bearers of such a power than a foetus’s whole body or brain (as on OBED). Strong emergence, however, doesn’t allow us to afford much weight to such intuitions.

This point can also be brought into focus from another direction. The biological path to a fully formed human is clear: the sperm fertilizes the egg, and the two become a zygote, which then goes on to develop, in normal cases, into an embryo, then a foetus, then an infant. The emergent dualist’s motivating considerations are that (1) the soul depends on the body such that the former can’t naturally pre-exist the latter; and (2) consciousness is irreducible, so that by the time a foetus is conscious, it must have a soul. But these considerations don’t settle when along the developmental path from conception up to the first conscious episode the soul emerges. Putting it right at fertilization and attributing it to the parents’ powers (through their gametes) turns out to be a version of traducianism.Footnote ⁵

But perhaps I am making short shrift of the worry about an intuitive mismatch. Doesn’t it make more sense that a functioning brain would give rise to a soul than that two gametes would? It seems to me that the answer to this depends on the perspective one assumes, and it isn’t easy to say which the right one is. The perspectives differ in their answers to two questions. First, do we think of souls in (broadly) Aristotelian or Cartesian terms? That is, are souls principles of all vital activities or only of the mental life? Some historical representatives saw traducianism as attractive because its negation was understood to entail that the foetus is initially not alive (Gregory of Nyssa Reference Schaff and Wace1893, On the making of man, 29.5). But even on a broadly Cartesian understanding – the one that I and most contemporary emergent dualists endorse – it matters whether we judge intuitive fit in terms of the child or of the parents. In terms of the life of the child, there may seem little reason to suspect a soul before there’s a rich, functioning biological basis to interact with it and to sustain one’s mental life. On the other hand, if the parents are mind–body composites, and we don’t assume that this is merely a stage in their life, then it is quite intuitively fitting that in reproduction they would bring forth creatures of the same sort; mind–body composites from the start (cf. Shedd Reference Shedd1888, 10–11, 19–22).Footnote ⁶

Traducianism for emergentists: objections

Thus far the initial case. Now, I turn to some potential objections to PGED, hoping further to elucidate the view in the course of answering them.

1. Better one body than two. First, there are important disanalogies between living bodies on the one hand and pairs of individuals engaged in reproduction on the other. The former (some might say) are substances or natural unities – arguably more natural candidates for being bases of emergent soul-creative powers than a mere accidental aggregate of a pair of potential parents.

The premise of this objection can be granted; living bodies are indeed more natural unities than (temporary) pairs of them. However, the details of PGED render this irrelevant. Powers are very often realized in interaction with other things endowed with other, fitting powers. On a certain level of description, this characterizes sexual reproduction itself: (fertile) males and females have the power to produce offspring upon gametic union (i.e., intercourse or its medically facilitated alternatives) with a suitable partner. Analogously, we needn’t attribute the emergent power to produce an offspring’s soul to the aggregate or mereological composite of the parents; it’s more plausible to attribute to each of them individually the power to produce an offspring’s soul upon gametic union with a suitable partner. This is very much in line with the spirit of Naturalness in that the power to produce the offspring’s body and the power to produce its soul are similarly allocated.

2. The man’s not there. The second problem concerns ‘locating’ the requisite power in time and space. Many attempts at reproduction fail to result in conception. Even if they succeed, there’s a temporal distance between intercourse and conception (in the human case, on the order of magnitude of days). If souls are only brought into existence when reproduction is successful, but the male may be gone by the time reproduction succeeds, how can the exercise of his power contribute to the emergence of a new soul?

Notice that this problem is not unique to the procreation of souls, as we normally attribute the power to procreate in general to males and females as complete organisms, despite the fact that males are not around at the moment of conception. The natural solution, already mentioned in the initial statement of the traducian model, is to identify the gametes as the vehicles or media for the organism’s exercise of its reproductive powers. The advantages here are obvious, for the sperm and the egg are both present right up until fertilization, and are moreover constitutive of the resulting zygote. Note that this solution applies equally to the procreation of both body and soul.

One might worry that bestowing such a role on the gametes is explanatorily inadequate. We understand how males contribute to the production of their offspring’s body by or through their sperm, but we can’t understand how they can contribute to the production of their offspring’s soul by the same means. Relatedly, the sperm and the egg may not seem uniquely suitable for mediating such soul-producing powers compared to other cells. Though they doubtless differ from other cells in their constitution and especially (and relatedly) their roles, these differences may not seem sufficiently explanatory of the different powers we’re attributing to them. However, the upshot of strong emergence is just that this sort of explanatoriness cannot be expected – embracing strong emergence requires tempering one’s demand for an explanation.

3. Is this traducianism? Third, engaging with a proposal very much like this one, Crisp wonders whether it’s truly a version of traducianism:

Crisp here seems to take the differentiating factor between creationism and traducianism to be whether the new soul comes into being ex nihilo or as constitutively dependent on the parents’ souls. Accordingly, by Crisp’s lights, traducianism also entails that souls aren’t simple; by ‘soul-stuff’ he means ‘an immaterial equivalent to sub-atomic particles’ (Crisp Reference Crisp2006, §36). I’ve argued against these claims in the section on ‘What is traducianism?’; constitutive dependence (and, with it, composition) is missing from multiple canonical descriptions of traducianism, and thus shouldn’t be considered essential to the view. Rather, the distinguishing feature between standard versions of creationism and traducianism is whether the new soul’s production requires divine intervention or is accounted for by the powers of the parents.

An interesting question remains: are the account Crisp alludes to and the one I proposed in the previous section versions of emergent dualism? I submit the answer depends on whether the gametes’ power to ‘generate soul-stuff’ supervenes on their physical and biological properties, or whether their having it is somehow dependent on, or explained by, their originators’ (having) souls. I’m inclined to say the latter – in which case the resultant view would arguably not be a version of emergent dualism – but my argument does not hinge on the answer.

Traducianism for creationists

Creationists believe that God creates the soul of every human being (and every soul-endowed animal) at the time of conception or at some point during gestation. Many historical luminaries have endorsed the view, such as Jerome, Avicenna, most medieval scholastics including Bonaventure and Aquinas, most Reformed theologians, and Descartes (Spruit Reference Spruit2014, 32, 39–49, 91–93, 111). Recent proponents include the early Swinburne (Reference Swinburne1997, 198–200),Footnote ⁷ Foster (Reference Foster and Corcoran2001, 28–29), Hershenov (Reference Hershenov2024), and qualifiedly, Farris (Reference Farris2018a, Reference Farris2018b). It is also taught in the Catechism of the Catholic Church (para. 366), promulgated in 1992. Creationism is no doubt a metaphysically heavy account, and one that is not amenable to most forms of ‘naturalism’. Still, it is attractive to many theists. Further, some consider it independently motivated by the radical discontinuity between mind and matter; it is hard to see how non-minded things can bring minds into existence.

Consider now the following version of creationism:

I’ll argue again, as I did in the section on ‘Traducianism for emergentists’, for two claims. First, that Immediate nomic creationism (henceforth: INC) is a version of traducianism; second, that it is no less plausible than alternative versions of creationism.

I don’t need to spend much time on the first point. Naturalness is secured by the model’s feature that God’s soul-creation counts as a law of nature. Parental dependence and Immediate ensoulment are, in turn, secured by stipulation – not implausibly, though, since nomological dependence is well-trodden ground (Jansson Reference Jansson2015; cf. Lewis Reference Lewis1973, 563–564).

The second claim – that of at-least-equal plausibility – requires much more argument. From the creationist’s perspective, INC makes two controversial claims. First, that soul-creation happens immediately at fertilization; second, that it counts as a law of nature. One might think that there’s a further commitment to parental dependence, but that seems to be entailed by the combination of the former two – it is very hard to see how lawful soul-creation immediately upon fertilization could avoid nomic dependence on the parents. The question, then, is whether the commitment to immediacy and lawfulness should seem costly to the creationist. Let us consider them in turn.

Immediacy. One might think that nothing prevents God from creating the soul at the time of fertilization – indeed, fertilization is arguably the least arbitrary occasion for ensoulment. However, this topic is wrapped up in a large debate about when organisms, lives, individuals, and persons begin, and whether the zygote is a suitable entity for ensoulment. Since this debate is not uniquely relevant to creationist versions of traducianism, but rather impinges on all versions of the latter, I shall postpone its discussion until the section on ‘Monozygotic twinning’.

Lawfulness. I shall present a two-step argument for the lawfulness of soul-creation not being a cost for the creationist. First, I’ll argue that the existence of psychophysical laws removes much of intuitive push against this idea; second, I’ll argue that a divine decree to create souls upon fertilization will count as a law of nature according to many theistic accounts thereof.

First, psychophysical laws. For explanatory reasons, substance dualists need to, and typically do, acknowledge laws of nature which govern psychophysical interactions. These are the regularities that specify the relations between the mental and the physical. The central motivation for introducing such laws is to explain the strikingly close correlation between the mental and the physical. A wide variety of physical substances can affect one’s mental states – moods, etc. – and all sorts of physical conditions can impair mental functioning – e.g., dementia. On the other hand, one’s beliefs, desires, and volitions influence one’s bodily behaviour all the time.

Examples of psychophysical laws would include specifications of interrelations between C-fibres firing and pain sensations, or firings in the visual cortex and visual experience. Further, though somewhat less clearly, they may also specify how certain mental contents play a role in bringing about actions and changes in mental states. Psychophysical laws have material macro-states on one side and immaterial (partly) qualitative states on the other. As such, they’re evidently not deducible from microphysical laws.

Thus, the creationist should accept (independently of our current topic) that there are laws of nature governing the interaction of matter with immaterial, directly divinely created entities. Once she accepts this, it won’t be easy to argue that there cannot be laws of nature governing the production of such immaterial entities.

Second, theistic accounts of the laws of nature. Suppose God decides to create and sustain the world such that every conception of the right kind of creature (and no other event) coincides with the creation of a soul linked to the body of the new foetus. Call this the creationist decree (CD). I shall argue that on prominent accounts of how God might be responsible for the laws of nature, CD may count among them.

Start with regularity theory or ‘Humeanism’ – such a view holds that the laws of nature are mere statements or descriptions of the regularities found in nature (Hildebrand Reference Hildebrand2023, 5–7). The simplest version of Humeanism, naive regularity theory, has the upshot that any universal regularity counts as a law. On such a view, CD shall count as a law.

However, other versions of Humeanism, such as best systems accounts, introduce restrictions as to which regularities might count as laws (though they all agree in not attributing to them any ‘governing’ or modal force) (Hildebrand Reference Hildebrand2023, 9–18). Might CD fail to be a law of nature on some of these accounts? Remember that here we are assuming substance dualism and the existence of psychophysical laws of mind–body interaction. This means that souls are among the entities the laws are supposed to cover; they’re part of the ‘Humean mosaic’. The acknowledgement of psychophysical laws also means that reference to immaterial souls cannot by itself exclude a regularity from lawhood. It’s hard to imagine, then, any extensionally adequate system of regularity-descriptions that doesn’t mention the coming into existence of souls, or one that doesn’t tie them to fertilization (or some event in its immediate vicinity). The upshot, again, is that it’s very plausible that if God decides to create and sustain a world with souls coming into existence upon all instances of fertilization and never otherwise, then regularity accounts will render CD lawful.

Foster (Reference Foster2004; cf. Plantinga Reference Plantinga, Bergmann and Brower2016) argues that not all regularities causally imposed by God on nature count as laws. For him, this ‘imposition’ has to happen in a particular way; to count as laws, the regularities must be imposed ‘as regularities’ (Foster Reference Foster2004, 157). By this, Foster seems to mean that regularity in its general form – that is, abstracted away from the particular instances which it describes – has independent explanatory relevance for how things turn out in the world, and, by implication, in counterfactual scenarios.Footnote ⁸ I see no barrier to God imposing CD ‘as a regularity’, such that his commitment is counterfactually robust and holds explanatory weight. Thus, on Foster’s ‘causal account’, again, CD would seem to count as a law of nature.

Consider now a theistic interpretation of the ‘relations between universals’ view. According to the theory proposed by Dretske, Tooley, and Armstrong – thus sometimes called DTA – laws of nature are nomic necessitation relations which hold contingently between natural-property grounding universals (e.g., Armstrong Reference Armstrong1983). That such relations hold and that their holding has the consequences characteristic of the laws of nature is usually thought to be brute, which grounds important objections to the theory (Lewis Reference Lewis1983, 366). However, an attractive theistic variation of the view might cash out the contingent necessitation relation holding between properties F and G as something like ‘God has decided to cause all Fs to be G’, or ‘God has decided to bring it about that all Fs are G’ (cf. Békefi ms.). Such an account, too, could easily classify CD as a law.

The argument of this section has contended that certain plausible forms of creationism are also forms of traducianism. This result might strike some as surprising or even suspicious, but it results from the combination of three defensible ideas: first, the sufficiency of lawfulness for Naturalness (cf. the section on ‘What is traducianism?’); second, psychophysical laws already accepted by the creationist; and third, distinctly modern accounts of God and the laws of nature (cf. Koperski Reference Koperski2025). The ensuing picture may differ from more traditional, ancient and medieval versions of traducianism – but that is not, in itself, an objection. The argument might, however, be vulnerable to other objections, especially since it does not cover all important theories of the laws of nature. I now turn to these objections.

Traducianism for creationists: objections

1. Other theories of laws. There do seem to be accounts of the laws of nature on which God’s direct creation of souls would not count as a law of nature. On a theistic version of primitivism about laws (cf. Maudlin Reference Maudlin2007, 14–21), God could exercise providence by creating primitive laws, including perhaps ones concerning soul-generation – but the governing effect of such a law would be metaphysically distinct from the direct creative acts of God (cf. Koperski Reference Koperski2025, S258). Something similar holds for theistic dispositionalism, the view that God creates natural laws by creating particular things essentially characterized by a fixed set of dispositions. The laws of nature would then possess their efficiency in a ‘bottom-up’ way, being grounded in the essential dispositions of particulars. On dispositionalism, laws of nature are expressions of these dispositions – descriptions of how particulars and their combinations interact (Adams Reference Adams2018; Page Reference Page2015).

Two questions crop up. First, could God create natural laws governing soul-generation on a primitivist or dispositionalist account of laws? Second, would such a view be implausible by the creationist’s lights? I address these two questions in turn.

Answering the first question from a primitivist perspective is easy. If laws of nature are primitive, then no constraint on their content would allow one to say that there can’t be any laws about such-and-such entities. There appears to be no principled reason for God not to be able to create a law governing soul-generation.

The answer for dispositionalism requires more argument. Start, again, from psychophysical laws. A dispositionalist account of psychophysical laws would very likely involve strong emergence and brute facts about composition. Psychophysical laws attribute powers to brains or nervous systems, or perhaps realizers of certain functions which cannot be explained by the powers of their constituents – higher-level powers that make it possible for certain bodies to interact with immaterial minds or souls. These laws would therefore require facts or laws about composition; that is, facts about what configurations of lower-level entities give rise to higher-level entities with emergent powers and properties. On strong emergence, these facts about composition cannot be laws of nature, since otherwise the composition laws plus the lower-level entities would entail the higher-level powers and properties. So, what remains is the option that these facts about composition are brute facts, and God has engineered the world making use of composition facts favourable towards interaction (Page Reference Page2020, 350). But if there are brute facts about composition ensuring that bodies or brains are endowed with soul-interacting powers, there may as well be brute facts about composition bestowing (partial) soul-producing powers on higher animals. These, then, would ground laws of soul-production or soul-generation.

Turning to the second question, it must be conceded that these primitivist and dispositionalist models of traducianism are not versions of creationism as the latter is typically understood. Otherwise, all patristic views surveyed in the section on ‘What is traducianism?’ would have to count as ‘creationism’ on account of their shared belief that God has created everything, including the powers they possess. Moreover, the dispositionalist model outlined here is hardly distinguishable from PGED (developed in the section on ‘Traducianism for emergentists’) – and it might be stretching the views too much if one model can simultaneously count as a version of creationism, emergentism and traducianism. Instead, call the traducian view consistent with primitivism and dispositionalism Theistic nomic non-creationism (TNNC):

That TNNC is not a version of creationism doesn’t mean that creationists would be justified in considering it implausible. For again, there is no principled reason why God couldn’t establish laws of nature governing soul-generation. One might object that such laws would be somewhat esoteric even relative to laws governing mind–body interaction – an extra commitment on the traducian’s part. However, this cost is plausibly outweighed by the benefits, since the lawfulness of traducianism involves significant cost reduction by eliminating regular divine intervention. Recall that, in contrast with the theories covered in the previous section, primitivism and dispositionalism cannot render divine action without secondary natural causes; therefore, primitivist and dispositionalist creationism involves postulating innumerable direct divine interventions throughout history. However much the traducian spends on laws of soul-generation, she saves at least as much by not postulating countless divine interventions. Thus, even though TNNC isn’t a version of creationism, it ought not seem implausible by the creationist’s lights.

2. Reject psychophysical laws. Much of the argument for the palatability of soul-generation’s being lawful is predicated on the necessity of psychophysical laws for any plausible account of substance dualism. Can’t this be called into question?

One way would be to embrace mind–body interaction without laws. This was Swinburne’s (Reference Swinburne1997, 186–196). But one would need to maintain that this lawless interaction exhibits patterns of extreme regularity, for this is severally evident. It is supported by (1) introspective observations of correlations between one’s own mind and body; (2) the success of much medical, psychiatric, and cognitive scientific research; and (3) the dialectical threat of other-minds scepticism, reliable mind–body correlation being a prerequisite of our knowledge of other minds. The ‘interaction without laws’ view must treat this regularity as an inexplicable mystery, as Swinburne conceded (Reference Swinburne1997, 195–196), which renders this view seriously unattractive. Curiously, Swinburne suggested that it is God who ‘brings it about that there is linked to the brain of an animal or man a soul which interacts with it in a regular and predictable way’ (Reference Swinburne1997, 198). But God plausibly does this in one of two ways: either by establishing psychophysical laws, or by directly intervening. The former establishes my point, while the latter involves innumerable miracles at every instant – again, not an attractive option.

Swinburne’s change of mind is instructive. His earlier objection was that while we can make reliable generalizations about the patterns of such interaction, there’s no prospect of unifying them into some simple and powerful theory, which is a prerequisite of them counting as laws (Swinburne Reference Swinburne1997, 186–196). However, in recent work, he seems to have moved away from this requirement, and endorses psychophysical laws – while at the same time expressing scepticism that these could ever be developed into a ‘general theory’ (Reference Swinburne2019, 159–160) – exactly under the pressure of the alternative’s mysterious implications:

The other option would be to deny mind–body interaction in either direction and adopt psychophysical parallelism instead. But psychophysical parallelism is deservedly out of fashion, being disastrously implausible, failing to explain (and indeed threatening to undermine our knowledge of) all mind–body correlation. Thus, I concede that the substance dualist who is neither an interactionist nor an epiphenomenalist is exempt from my argument; however, there might not be any philosopher of mind today to whom this exemption applies.

The first soul

So far, I have argued that both emergent and creationist dualism are hospitable to versions of traducianism. To complete my case for the plausibility of traducianism, I must consider two problems to which it is arguably uniquely vulnerable. These are the problem of the first soul and the problem of monozygotic twinning. This section deals with the former; the next section addresses the latter. Now, let us turn to the first soul.

Traducianism seems to have a disadvantage relative to emergentism and creationism in that its account of how souls come to be assumes that other souls already exist. Therefore, it seems like it, and it alone, cannot account for the existence of souls in toto. In this section, I shall argue that this impression is mistaken; traducianism is no worse off than its alternatives when it comes to explaining the origin of the first soul. To see this, we will have to consider the three models of traducianism that I have proposed over the course of this paper. These are (1) PGED (see ‘Traducianism for emergentists’), (2) INC (see ‘Traducianism for creationists’), and (3) TNNC (see ‘Traducianism for creationists: objections’).

Start with PGED and consider what emergent dualists would say about the origin of the first soul. (Recall from the end of the section on ‘What is traducianism?’ that we are restricting ourselves to sexually reproducing species.) They would have to say that once upon a time, a soulless animal gave birth to an offspring that differed from it in just the right ways to possess and exercise this new power to produce a soul for itself. I submit that traducians of the relevant sort can say pretty much the same thing. To wit: once upon a time, a soulless animal gave birth to an offspring that differed from it in just the right ways to possess and exercise this new power to produce a soul for its offspring upon reproduction. Or, perhaps, the first soul came about only when a suitable pair of animals, each endowed with such a power, first mated successfully.

Consider now INC. Here, the traducian should agree with his non-traducian counterpart: the first soul, just as the later souls, was created by God. It’s just that, in contrast with the non-traducian counterpart, the traducian maintains that God creates souls in a way that depends nomologically upon the parents. Given such nomological dependence, the first soul could have come about in one of two ways. Either, like with PGED, a soulless couple fulfilled the nomological antecedents of soul-production, and brought about offspring that God then endowed with a soul; or INC theorists could deny the natural continuity of soulless and soul-endowed animals, and maintain that God directly created the first couple of soul-endowed animals. The nomological antecedents of soul-production, then, would all involve the reproductive acts of animals that already have souls.

Finally, TNNC. Here, it seems to me, either of the previous two accounts could be adopted. One might hold with the naturalists that the first soul arose naturally when some animals naturally developed the emergent power to beget it; alternatively, one might take one of the theistic options and say that the first soul, or the first soul-endowed animal, was created directly by God. All of these seem palatable, and the present dialectic doesn’t require us to find the best one among them.

In sum, then, each of the traducian models does at least as well as their non-traducian counterparts in explaining where the first soul might have come from.

Monozygotic twinning

Up until around two weeks of gestation, a zygote might split into a pair of monozygotic twins. This grounds two thorny objections to traducianism. The first objection is that traducianism entails conceptionism – the view that the life of the organism or the existence of the individual begins at conception – but conceptionism is implausible in the light of monozygotic twinning because, arguably, no human individual can become two, so the zygote can’t be a human individual.Footnote ⁹ The second objection is that even if conceptionism is consistent with monozygotic twinning, Parental dependence and Immediate ensoulment aren’t, since if both twins have souls, then at least one of them did not originate from the parents at conception.

The first objection is a wash. Granted, the falsity of conceptionism strongly disfavours traducianism, since traducianism without conceptionism is rather implausible.Footnote ¹⁰ On the other hand, the truth of conceptionism favours traducianism, since it is a ‘prediction’ thereof, while non-traducian versions of emergentism and creationism are neutral with respect to it. Therefore, reasons against conceptionism (such as the phenomenon of monozygotic twinning) are reasons against traducianism – but reasons for conceptionism are reasons for traducianism! And there are indeed independent arguments for the human organism’s or person’s beginning to exist at its conception (e.g., Colgrove Reference Colgrove2026; Jacobs Reference Jacobs2021; Lee Reference Lee2004; Pruss Reference Pruss and Napier2011). Thus, the plausibility of traducianism is sensitive to the plausibility of conceptionism, but this is not in itself an objection unless one considers conceptionism unlikely.

The second objection requires considering traducian interpretations of monozygotic twinning. Simplifying Skrzypek’s (Reference Skrzypek2024) thorough taxonomy, the main views in general are ‘fission’ (the original organism ceases to exist and two new come into existence from it); ‘budding’ (one organism is produced by the other); and ‘original overlap’ (both organisms exist from conception, but they share a body or overlap completely). For our purposes, fission and budding can be treated together. In the following, I consider the plausibility of the ensuing two traducian models.

1. Fission and budding as asexual reproduction. Monozygotic twinning may be taken to be a kind of asexual reproduction (cf. Pruss Reference Pruss2013, 379, 429n37, 441n3). One could then either say that twinning is beyond the purview of traducianism in same way that asexually reproducing species are, since traducianism is only meant to account for souls coming into existence through sexual reproduction (cf. the section on ‘What is traducianism?’); alternatively, one might argue that Parental dependence can still be met in monozygotic twinning if, given asexual reproduction, one understands ‘parent’ to refer to the original zygote.

The costs here are the following. First, one might argue that traducianism is supposed to account for at least all human souls (except, perhaps, the first one), and excluding some human twins from the account – or at least significantly revising the account to accommodate them – is a retreat. This is fair enough, though mitigated by the strong possibility that asexually reproducing species already require some non-traducian account. Second, it is typically thought that twinning involves fission or budding, but some have argued that attributing fission or budding to souls entails that souls are not simple substances – contrary to popular substance dualist views (Crisp Reference Crisp2006, §§35–49; Farris Reference Farris2015, 381–385, Reference Farris2016, 45–49). But we ought not assume that if the biological process is one of fissioning or budding, the soul must undergo the same. Instead, it might be that the fissioning or budding zygote has the power, strongly emergent or nomically bestowed, to produce one (budding) or two (fission) new souls in twinning. This does not require that souls be composed of parts.

2. Original overlap as double ensoulment. The other option is to say that when a zygote splits into two in twinning, the souls of the ensuing two zygotes were there from the beginning (Koch-Hershenov Reference Koch-Hershenov2006, 158–160). The central challenge here is to explain why there were two souls in the beginning without falling into a view that allows for the generation of multiple souls at every fertilization. The solution must consist of two steps. First, the traducian must say that some gametes, or all gametes in some circumstances or with some probability, have the power to produce two souls upon fertilization. Second, he must propose some mechanism for ensuring that two-soul fertilizations are coextensive with fertilizations that lead to monozygotic twinning (unless terminated). Here, again, I see two options. Either it is the presence of the two souls that brings about or explains twinning through some special instance of mind–body interaction; or, two-soul fertilizations must involve some biological difference in the zygote which is both necessary and sufficient for monozygotic twinning (bracketing termination). If twinning follows deterministically from the initial conditions at fertilization, then the latter option is tenable; if there’s an indeterministic factor, then the former option cannot be ruled out.

Here, the costs are harder to gauge. Only a necessary or sufficient, initially absent environmental factor could falsify both mechanisms. The current science regarding what determines whether monozygotic twinning happens is controversial and unclear, but such a factor is not in sight (Herranz Reference Herranz2015; Jin et al Reference Jin, Jan and Zenker2024). Even so, one might argue that making pre-existence palatable comes at the cost of additional claims (e.g., double-soul generating gametes) that detract from traducianism’s parsimony. In a sense, this is true. However, it is not surprising in that once one ties the origin of souls to procreation, variations in procreation will require introducing complexity into the account.

Let us take stock. Monozygotic twinning is a double whammy against traducianism, counting against it twice by questioning both conceptionism and Parental dependence. Some of this is offset by positive arguments for conceptionism and traducianism; if one is already a conceptionist, then the implausibilities of traducian models of twinning might be compensated for by the fact that traducianism predicts conceptionism. However, for the neutral observer, monozygotic twinning does seem to count, at least somewhat, against traducianism.