Spinoza’s Evanescent Self

  • Sanja Särman orcid logo (Hong Kong University)


Selfhood is a topic of great interest in early modern philosophy. In this essay, I will discuss Spinoza’s radical position on the topic of selfhood. Whereas for Descartes and Leibniz, there is a manifold of thinking substances, for Spinoza, there is, crucially only one: God (1p14; 2p1). Minds, for Spinoza, do not have substantial status, they are instead merely complexes of ideas (2p15), and thus complex modes of the one substance: God. Observations such as these often lead Spinoza’s readers to the conclusion that, whereas for Descartes as for Leibniz, human beings have robust or genuine selves, this is not so for Spinoza. However, this reductionist interpretation is also challenged—in recent times most intriguingly by Koistinen (2009). Koistinen has argued that there are, fundamentally, human selves of whom agency can be predicated in Spinozism. In this paper I discuss to what extent this is true. In section 1, I introduce the reductionist interpretation of selfhood in Spinoza’s thought. In section 2.1, I present and criticize Koistinen’s proposal. In section 2.2, I acknowledge the strength of Koistinen’s view that insofar as human beings act, they are God somehow. In section 3, I propose an alternative reading of human selfhood in terms of witnessing being acted out rather than in terms of being an agent. This view isprima facieparadoxical. In section 4, I nonetheless support it by highlighting that Spinoza seems to have seen practical benefits in knowing oneself to be acted out by God. I conclude the essay by pointing out some comparative directions for future research.

Keywords: Spinoza, selfhood, agency, phenomenology

How to Cite:

Särman, S., (2022) “Spinoza’s Evanescent Self”, Journal of Modern Philosophy 4: 5. doi:



Published on
18 Feb 2022
Peer Reviewed


Selfhood is a topic of great interest in early modern philosophy. Indeed, ‘the modern subject’ or ‘the modern self’ is often considered to be born in the late early modern period.1 In this essay, I will discuss Spinoza’s radical position on the topic of selfhood. Whereas for Descartes and Leibniz, there is a manifold of thinking substances, for Spinoza, there is only one: God (1p142; 2p1). Minds, for Spinoza, do not have substantial status, they are instead merely complexes of ideas (2p15), and thus complex modes of the one substance: God. Observations such as these often lead Spinoza’s readers to the conclusion that, whereas for Descartes as for Leibniz, human beings have robust or genuine selves, this is not so for Spinoza.3 However, this reductionist interpretation is also challenged—in recent times most intriguingly by Koistinen (2009). Koistinen has argued that there is, fundamentally, human selves of whom agency can be predicated in Spinozism.4 In this essay I discuss to what extent this is true. I point out some problems with Koistinen’s view but also highlight its main strength. More precisely, I grant that insofar as we act we are God somehow, but I dispute that insofar as we are God somehow we are ourselves. In short, I do not believe that Spinoza’s texts support attributing the source of human agency in God to anything like a human self.

The essay is structured as follows. In section 1.1, I introduce the bundle view, a reductionist interpretation of selfhood in Spinoza’s thought. In section 1.2, I present exegetical reasons why the bundle view remains viable in the face of two objections, namely, the objection from Spinoza’s ample use of the reflexive pronouns (1.2.1) and the objection that Spinoza’s thesis that we are always aware of having true ideas when we do so, presupposes the existence of a self (1.2.2). In section 2.1, I present and criticize Koistinen’s proposal. In section 2.2, I acknowledge the strength of Koistinen’s view that insofar as human beings act, they are God somehow. In section 3, I suggest that if selfhood can be interpreted in terms of witnessing rather than acting, there may still be room for human selfhood in Spinozism. In section 4, I partially support this view by appealing to the practical benefits Spinoza seems to have seen in the idea that we are not fundamentally the thinkers of our thoughts (or the doers of our deeds). I conclude the essay by briefly pointing out a comparative direction for future research.

1. In Defence of a Reductionist View of Self in Spinozism

1.1. Support for the bundle view

The notion of a self is difficult to expound. Let me tentatively adopt a standard (albeit vague) understanding of selfhood. According to this understanding, the self meets the following two criteria. First, a self serves to unify the experiences, thoughts or acts of someone. It is worth noting that, although we could stipulatively define a self as that to whom different kinds of phenomena belong (e.g., experiences, thoughts or acts), agency often takes pride of place in Western accounts of selfhood. That is, a self is often considered the centre of the actions deliberated and carried out by some individual, rather than, say, the impressions or passions an individual experiences. But what is the feature in virtue of which acts belong to a self? One answer is that the self is not only the subjective centre of its actions, but also their source. When the self is the only source of its actions, there is Agent Causation.5 We can formulate this selfhood criterion for agents as follows: it is by belonging to the self of an agent (thinker) that a series of acts (thoughts) belong to the same agent (thinker).

The reductionist reading is supported (but not vindicated) by the fact that selfhood appears to serve no such unificatory role in Spinoza’s philosophy. For Spinoza exclusively appeals to alternative means of individuation for modes such as human beings.

Spinoza nowhere discusses persons or selves. Instead he speaks of modes, minds and bodies. The human being is a mode. It can be conceived now under the attribute of thought, as a mind, and now under the attribute of extension, as a body. The two attributes are linked, for the human mind is the idea of its body (2p11). This intentionality notwithstanding, Spinoza famously holds both that the body and the mind are numerically identical (2p7s), and that physicality and mentality are each causally closed (2p6). (That they are causally closed means that a mental effect will never have a physical cause, and vice versa.) Moreover, the mental and physical orders are causally and structurally isomorphic. As a consequence, the physical relations that build and disintegrate bodies (relations Spinoza thinks of in terms of ratios between motion and rest) have mental counterparts. Just like our bodies are aggregates of physical parts (2post1), so minds must be aggregates of mental parts. These mental parts are ideas. A mind, it appears, is a composite idea and its component parts are ideas (see 2p156). What, on the mental side, explains that a certain collection of ideas is one mind? Is the unification of complexes of ideas into minds to be explained by a self?

Spinoza makes no appeal to a self in order to explain the unity of minds. Instead, Spinoza’s understanding of individuation—which applies to minds as well as bodies—appears functional or structuralist. An individual is functionally defined when it counts as the same individual so long as it can continue to carry out the same function, for example, some action (as in the Definition of the ‘Physical Excursion’ of his Ethics), and structurally defined when its individuality is constituted by the structure that obtains between its parts, rather than by the parts themselves (2lem4; 4p39d and s).7 It appears reasonable to take Spinoza’s preference for functionalist and structuralist explanations to imply that, within the confines of Spinozism, selfhood becomes redundant, at least with respect to individuation. For there is no need to posit a self to unify minds when functional and/or structuralist explanations suffice to individuate and unify entities.

But can Spinoza really maintain that there is nothing to minds but ideas?

According to the bundle theory of the self, there is nothing more to the mind, which is a mode of God, than a bundle of ideas that is merely structurally and/or functionally unified. But if there is nothing to the mind but its ideas, then with what right can I say that I have these ideas? Let us grant, for now, that it is possible for ideas to form a dynamical pattern whose function or structure singles me out as myself. This does not explain, however, why the ideas that form this pattern would ever qualify as my acts or, to speak from the mental point of view, as ideas I think. Why is it, then, that we say that we think or act, instead of saying that ideas and their concomitant volitions conglomerate in ways that constitute what we are? The structuralist or functional account of mental unity appear to provide no answer as to why the activity of thinking appears to be carried out by me, rather than merely individuating the complex entity that I am. More clearly put, structuralist or functional accounts of mental unity seem not to do justice to the phenomenon of agency. I will return to the objection from agency in section 3. Before doing so, allow me to first address two related objections against the reductionist view.

1.2. Active ideas

One might think that if Spinoza did not think that we ourselves were agents or thinkers, he should not have said so. ‘But,’ an objector to the reductionist position now contends, ‘he does.’ According to this objector, activity is predicated of selves and not merely of ideas at numerous passages in Spinoza’s works. I cannot list all passages that are candidates for being included on the objector’s list here. But consider, for example, the definition of bondage in 4pref. Here, Spinoza states that ‘the man who is subject to affects is under the control, not of himself, but of fortune, in whose power he so greatly is that often, though he sees the better for himself, he is still forced to follow the worse’ (emphasis added). If the reflexive pronoun refers to the man’s self and not simply to the man, then this testifies to the existence of (human) selves in Spinozism. More seriously, in the cited passage, the human being appears to be conceptualized as a subject which underlies the affects it undergoes and the ideas it has, and not as an entity constituted by the dynamical pattern these affects and ideas form. In a similar fashion, activity appears to be predicated of selves and not merely of ideas in 4p28s.

Passages like the one cited, the objector to the reductionist position contends, are not compatible with the doctrine that the sole activity that minds possess is the one that accrues to their ideas. Instead, it supports the view that (at least some) activity belongs to human selves and that this is what unifies the mind in question.

I grant that Spinoza does not deny that we are aware of ourselves and perceive ourselves as agents and persons. However, this does not mean that metaphysically, there are human selves. It is at least conceivable for there to be the phenomenological appearance of a self—which we can call the person—without there being any metaphysical entity such as the self. It is, differently put, possible for Spinoza to be an error theorist vis-à-vis the self.

Indeed, there appears to be reason to doubt that within Spinozism there is any sharp ontological distinction to be made between us as thinkers and the thoughts we think. The thoughts we think are ideas. According to a popular picture of what happens when we entertain an idea, we, as mental agents, consider it, ponder it, and reject it or accept it (for example).8 If that description is true, then we have solid reason to believe that there is a difference between us, the thinkers—the subjects—and our ideas—the objects of our contemplation. But Spinoza denies this picture. In his own words, he denies that ideas are ‘mute pictures on a panel’ (2p49s[II]). In contemporary terms, one could say that, for Spinoza, volitions supervene on ideas. That is, any attitude (whether pro or con) that one adopts vis-à-vis an idea is as it were accomplished by the idea itself. Since whether we affirm or negate an idea depends completely on the idea itself, one could call ideas auto-affirming (or auto-denying). There is then no need for a conceiver or a thinker to execute any act of thinking.9 (In the preceding sense, ‘thinking’ is intended to cover all mental activity. The use of ‘thinking’ [penser; cogitare] in this wide sense has a clear predecessor in Descartes, with the notable difference that for Descartes, unlike for Spinoza, mental life had to be conscious.)

As always when there is a collision of claims it is wise to consider which is more fundamental. Is Spinoza’s casual references to personal and reflexive pronouns more fundamental—or, more broadly, are those passages in which he speaks as if we are thinking and acting subjects more fundamental than 2p49? Or are we instead to treat as more fundamental his explicit rejection of a certain understanding of ideas as being merely the objects we contemplate and not the very activity of contemplation? Spinoza’s Ethics contains no independent discussion of what we are to make of his grammatically (and phenomenologically) correct use of pronouns. But he does extensively caution the reader to not mistake his or her ideas for the objects of thinking. Ideas are instead themselves active, in the sense that they themselves perform acts of conceiving and willing. In the scholium of 2p49 he draws important ethical consequences from this doctrine, showing how accepting that one is not the subject of one’s thoughts is in the end conducive to various kinds of practical progresses. And this move is by no means unique to the Ethics: the Short Treatise contains a chapter devoted to the same topic (KV, II, 18). By contrast, Spinoza does not explicitly discuss any practical consequences of us being selves. Spinoza thus appears to treat his theory of ideas as of greater systematic, and ethical, importance.

We should also bear with him relapsing into common ways of speaking. For the idea Spinoza is attempting to articulate—the idea that we stand in the same relation to God as it appears that our acts and states stand to us—strays so far from ordinary ways of thinking that he is sometimes compelled to recourse to linguistic conventions. We should not discredit one of Spinoza’s theses—the notion that since our ideas (including our perceptions, feelings, cognitions, volitions) are active, there is no need to posit ourselves as thinking (perceiving, feeling, cognizing, willing) subjects to account for that activity—merely because he sometimes relapses into conventional language. All original thinkers must occasionally do so, since our language is not suited to express their ideas. In sum, Spinoza’s conventional use of the pronouns does not entail that he subscribes to a self. There remains, however, a more serious objection to reductionism, namely the objection from self-awareness. How can we be aware of ourselves if there is nothing there for us to be aware of?

1.3. Self-aware ideas

There are two appeals to self-awareness that appear to challenge reductionist readings of the Spinozist self. The first one arises in connection with Spinoza’s theory of the evidentiality of truth.

In having true ideas, we are aware of having such ideas. Spinoza states this in 2p43: ‘He who has a true idea at the same time knows that he has a true idea.’ As an epistemological thesis, Spinoza’s claim is almost certainly false.10 I will not discuss its plausibility here, however, but instead comment on whether it undermines Spinoza’s arguably reductionist understanding of selfhood. It has been argued that in claiming that self-reflexivity necessarily accompanies truth, Spinoza is committed to the view that there is, in addition to the complexes of ideas—that is, minds, a self that can, in effect, be aware of having true ideas when it does so. Koistinen (2009: 164), for instance, writes that the demonstration of 2p43 shows that ‘the subject herself is part of the content of the second-order idea’.11 Since, moreover, that idea is adequate (being true), the subject herself must be adequately cognized. Hence, that subject must exist in its own right, distinct from the ideas that are being thought. The third person pronoun (‘he’) in 2p43 refers not to any idea, no matter how composite, but to the one thinking the idea: the self. This is no casual use of pronouns. This is no accommodation of the experiential reality of personhood. It is an explicit reference to the existence of an object of knowledge—the self—that must be different from the ideas it entertains: or so the objection goes.

I do not see that this objection works. As we saw above, Spinoza opposes a view according to which ideas are ‘mute pictures on a panel’. On Spinoza’s alternative view, ideas are active: they are activities of cognition and volition. It is important to appreciate the full import of Spinoza’s position. If ideas are acts of cognition, what stops them from also being acts of self-cognition?

Nothing stops them, the objector grants. But yet, she or he continues, if ideas are acts of self-cognition there is no reason why we ourselves should be aware of each self-aware idea. For each idea merely cognizes itself. But in 2p43 Spinoza claims that someone (‘he’) is himself aware of having the true idea, not that the true idea is aware of itself. There is no reason why the self-reflexivity of ideas should translate into the self-reflexivity of selves. Therefore, Spinoza is speaking only of the self-reflexivity of thinkers, not the one of thoughts, in 2p43.

The objection is strong. But I believe that Spinoza’s system furnishes us with the means to meet the pivotal claim that there is no reason why the self-reflexivity of ideas should translate into the self-reflexivity of selves. In order to show that there is indeed such a reason, it is necessary to use a Spinozist trick of the metaphysical trade: parallelism.

In the secondary literature, Spinoza’s parallelism is often referred to in the singular. Yet, as Melamed (2013: 152) helpfully explains, Spinoza espouses not one but two doctrines of parallelism. According to the first parallelism, the idea of a body is numerically identical with the body in question, but they are conceived through different attributes. This parallelism is the object of much scholarly debate. According to the second parallelism, any idea of an idea is similarly identical with the first-order idea (2p21). Spinoza explicitly states that: ‘[T]he idea of the Mind and the Mind itself are one and the same thing, which is conceived under one and the same attribute, viz. Thought’ (2p21s). The mind is thus the idea of the body (first parallelism). This mind is also identical with the idea of the idea that it is (second parallelism). But the idea of the body is complex (2p15). The idea of the mind is therefore an idea of this complex entity. Differently put, it is a comprehensive idea of all component ideas that together constitute the mind. Since Spinoza maintains that true ideas are self-cognizant, this picture may be conveyed in these terms: the idea of the mind is an act of self-cognition. What is cognized in this self-cognition are all individual acts of cognition and self-cognition that together constitute the mind. There is no need for a self in addition to the mind and the self-aware idea of the mind.

More precisely, there is no need for a self to be posited in order for a mind to be able to have ‘reflection, and consider itself as itself, the same thinking thing, in different times and places’ (these are Lockean features of personhood).12 A mind, somehow identical with a self-aware idea of said mind, suffices. If we understand persons as the self-conscious beings with whom we identify, we can, for the purposes of exploring the fate of selfhood in Spinozism, replace all talk of persons with talk of minds. The fact that we, as minds or as persons, can perform a synthesis of our experiences does not testify to the existence of an independently existing self.

Another, for the reductionist truly alarming, occurrence of a pronoun that appears to posit a self is 5p23s, where Spinoza states that ‘we feel and know by experience that we are eternal’. This is a notoriously difficult statement which, no matter how it is interpreted, appears to contradict much of what Spinoza says elsewhere. In addition, the statement appears to support the non-reductionist view that there is something essential to the person, the continued existence of which will guarantee the continued existence of the person. However, in accordance with what I have stated above, 5p23s can be construed so that ‘we’ refers to the meta-idea previously discussed, that is to the idea of the complex idea that constitutes our minds. The intended referent need not be a self in any way that we would ordinarily understand the term.

2.1. Koistinen’s proposal challenged

For the reasons stated above, the bundle theory is the theory of mind to which the Spinozist texts seem to lend themselves. According to this view, the self has no role to play in Spinoza’s theory of mind, since the mind is reduced to a bundle of ideas whose unity is functionally or structurally established. The unity of the mind is thus not explained in terms of the ownership of the self of all the phenomena arising in the mind. In contradistinction to the bundle theory, Koistinen has (2009) advanced the view that there are for Spinoza genuine or robust selves. Before acknowledging its main strength, I would like to point out two problems with Koistinen’s proposal. As far as I can see, these problems prove fatal to the proposal.

The question that Koistinen (2009) sets out to answer is ‘why [the] compositional mental entities that [by virtue of mind-body parallelism] correspond to bodies should be minds of subjects, i.e., why they couldn’t just be free-floating complexes of ideas’ (2009: 152, emphases added).

On Koistinen’s view, the bundle theory can afford no explanation as to why the minds that correspond to bodies should belong to subjects or selves, and is on this account to be rejected. Two questions immediately arise, questions related to each of the words emphasized in the citation. First, what is a subject or a self (for the sake of clarity, I’ll restrict myself to use the latter term in this sense)? I will return to this question at the end of this section. Second, why should minds belong to these selves? Is it the case that, unless selves are posited, we can’t make sense of the unity of our mental lives? Is the claim thus that unless such selves are posited, there is a brute fact about that unity? If that is so, Koistinen should have offered us a definition of the self that explains the unity in question. Let us therefore see how how Koistinen defines a Spinozist self.

Words like ‘person’, ‘self’ and ‘subject’ are often considered more or less synonymous. On a widespread theory, memory or some form of introspectively established psychological continuity is what accounts for the unity of the person.13 But for Spinoza, memory is carried out by the imagination and the imagination is the ‘faculty’ responsible for all one’s inadequate ideas, and inadequate ideas do not survive death. Memory, so to speak, belongs to one’s mortal coil and yet it is crucial, on Koistinen’s interpretation, that the self be immortal.14 It is important to realize that the fact that memory belongs to the imagination rules out any introspectively established continuity from being what the identity of the eternal self can consist in. Therefore, Koistinen makes an incisive distinction between the person and the self. The former alone is characterized by memory and imagination (see Koistinen 2009: 169). It transpires that the former alone is the subject of any form of introspectively established psychological continuity.

It is important to realize the full import of this position. If the self is the enduring center of an individual or a person, it is tempting to think of the self as the most or perhaps the only necessary ‘part’ or aspect of the person. But on Spinoza’s view, the continued existence of the self does not guarantee the continued existence of the person. The continued existence of the person might well still be memory or psychological continuity: this, however, is a matter unrelated to the self’s continued existence. One arguably essential aspect of a person is her memory, introspective continuity, or awareness of persisting through time.15 But as Koistinen rightly observes, all of this is but imaginary, for Spinoza. It follows that the continued existence of the self does not guarantee the continued existence of the person I think I am, that is, the intended referent of expressions such as ‘I’ or ‘myself’. For the continued existence of the person I think I am would presumably require subjectively experienced personal continuity.

The first problem of Koistinen’s proposal is thus the following. If the continued existence of the self (whatever it is) does not guarantee the continued existence of the introspectively accessible person, it is difficult to explain why one would choose to call the entity that will persist beyond my own destruction a self in the first place. Mustn’t there be some link between the persons we perceive ourselves to be and the selves we fundamentally are in order for the selves to be our selves in any relevant sense? One way of realizing the pertinence of the question is by attempting to identify the intended referent of ‘our’ above. Can something be my self without standing in any form of relation to what I perceive myself to be? Nonetheless, I agree with Koistinen that the enduring part of a mind could not be imaginary.

Having clarified what a Spinozist self is not, indeed could not be, let us attempt to clarify also what such a self is. As we saw above, it is not clear that it makes sense to speak of a self as entirely divested from the person. But perhaps Koistinen—who is in the business of finding a place for a robust selfhood in Spinoza’s philosophy—has a convincing proposal for what we should understand by this word in this context.

Koistinen formulates his positive understanding of the Spinozist self as follows: it is ‘identical with God insofar as he is conceived to act through a determinate force’ (2009: 160). More precisely, Koistinen defines a self—our selves—as follows: ‘I am God insofar as God is considered as acting through the force that individuates me and you are God insofar as God is conceived to act through the force that individuates you’ (2009: 159: Emphasis in the original).

It is worthwhile to consider how Koistinen’s view about the self in Spinozism is related to Spinoza’s own definition of the mind. We are first to recall that the mind, for Spinoza, is a (composite) thing, and so in defining what a thing is, he is also describing what a mind is at a general, attribute-neutral, level. Ideas are things, on Spinoza’s view, and the mind is a kind of (composite) idea.

Importantly, the definition I cited from Koistinen’s paper repeats almost verbatim Spinoza’s own definition of a particular thing. In 1p25c, Spinoza states that ‘particular things are nothing but affections of God’s attributes, or modes by which God’s attributes are expressed in a certain and determinate way’. Thus, a body is ‘a mode that in a certain and determinate way expresses God’s essence insofar as he is considered as an extended thing’ (2d1) and a mind is similarly a mode that in a certain and determinate way expresses God’s essence, but insofar as God is now considered as a thinking thing.

Koistinen’s view of the Spinozist self emerges as a result of two interpretative moves. First, he considers the relation x expresses the essence of y to mean that x acts through the essential force of y. We can call this the Expression Formula.16 I will grant Koistinen the Expression Formula, although I cannot undertake to defend it in this essay.17 Second, however, Koistinen transposes, as it were, the self over the thing which Spinoza defines in passages such as 1p25c, 2d1 and 3p6d. That is to say that Koistinen recycles Spinoza’s definition of thinghood for his own definition of selfhood. More clearly put, he defines self in the same way as Spinoza defines thing.

In the passages just mentioned, Spinoza consistently treats a singular thing—regardless of under which attribute it is presently considered—as something ‘that expresses, in a certain and determinate way, God’s power’ (3p6d). (Since God’s power is the same as his essence [1p36], things hereby express God’s essence.) Hence, a thing is something that acts through the essential force of God.

In contradistinction, Koistinen’s project is to root the mind Spinoza defines as a thing acting through the essential force of God when conceived as a thinking thing in a self. Yet, in order to determine that self more exactly, he borrows Spinoza’s definition of a mind. It is thus to be noted that Koistinen has not proposed an independent or illuminating account of what a self is. Instead, he has borrowed Spinoza’s definition of a mind, or a mental thing, and recycled it for the self. But if a self is distinct from a mind—if a self is that to which a mind belongs—then we would expect its definition to be different from that of a mind.

In order to see the problem more clearly, consider that any possible view on selfhood belongs to either of two basic categories. Either there is something, a self, that is necessary to the mind. Or there is no such thing, perhaps because the mind is simply an aggregate constituted by all of its parts, none of which is alone the one necessary feature of the mind, and the unity of this aggregate is to be functionally or structurally explained.18 As already mentioned, Spinoza appears to endorse the second view. Koistinen disputes that these appearances reflect reality. By appealing to a self, Koistinen attempts to claim that there is, for Spinoza, something which human beings are over and above the bundles of ideas that constitute each of their minds. For explanatory reasons, he is therefore not entitled to avail himself of Spinoza’s notion of a (mental) thing in explaining what the self is. But this is exactly what he does: for his understanding of a Spinozist self, cited above, and Spinoza’s own understanding of a thing do not differ, and the mind is simply a mental thing. All things are for Spinoza mental and physical at once, so in using Spinoza’s definition of a thing for the self, Koistinen actually postulates a mind (when that thing is conceived under the attribute of thought). If one grants that one is not allowed to recycle Spinoza’s definition of a mind in order to define that in whose existence the unity of the mind is to be grounded, Koistinen’s proposal must be considered as problematic.

Koistinen states that the minds should belong to selves and not be free-floating clusters of ideas, individuated structurally through their ratio of motion and rest. As we have seen, the should that Koistinen speaks of cannot be explanatory. For on his view of what a self is, the definition of self does nil to explain the unity of the mind—on the contrary, it accomplishes just as much (or, if you prefer, just as little) as Spinoza’s definition of the mind, for, on the assumption of the Expression Formula above, it is identical with that very definition. One major objection to Koistinen’s project is thus the problems that arise when one transposes a highly technical metaphysical definition from one entity to another and claims that one can thereby succeed in grounding the former in the latter.

2.2. The problem of agency

One possible motivation for providing Spinoza with selves is to explain how human action is possible in Spinozism. Koistinen’s case against the bundle theory is rooted in the difficulty to ‘reconcile’ said theory ‘with the possibility of acting’ (Koistinen 2009: 153). In the defence of the bundle theory provided above, I suggested that since ideas are active there is no need to posit selves as agents over and above ideas (and minds) in order to explain mental activity. The problem, however, is how to reconcile Spinoza’s theory of mind with acting as it is defined in his Ethics.19 In this work, action is understood in terms of adequate causation (3d2): human beings act when they cause effects that can be understood through human nature alone.

Koistinen rightly points out that if human nature is to be the adequate cause of anything whatsoever, then this nature must be somehow identical with God (Koistinen 2009: 153).20 For God alone is self-conceived and everything else is in God and conceived through God (1p15; cf. 1a4). Koistinen’s proposal—that in order to be agents in the first place, humans must ‘somehow’ be God—is well-supported by the texts: in 2p11c, for example, Spinoza claims that for a human being to think some idea is for God to think that very idea, ‘not insofar as he is infinite, but insofar as he is explained through the nature of the human Mind, or insofar as he constitutes the essence of the human mind, has this or that idea’. There is no simple mode-agency: all mode-agency is substance-agency, modified in some relevant ways.

However, I believe that Koistinen draws the wrong conclusions about selfhood in Spinozism from this premise. For Koistinen concludes that since human beings can act, human selfhood, qua the source of action, must be ‘divine’. While I grant that the source of human action is divine, I still maintain does not necessarily mean that something like a human self is ultimately the subject of action. Although Koistinen appears to be right that human beings cannot be agents unless they are somehow God, I do not agree with the consequences he draws from this premise: namely, that there are human selves, the subjects of human acts, which are somehow God’s essence. For there seem to be other, equally viable, ways of conceptualizing the situation. In being God somehow, human beings are perhaps not the agent-selves we so often think ourselves to be.

In what follows, I will leave the topic of agency in order to attempt to suggest that another kind of selfhood is nonetheless compatible with Spinozism.

3. Witness-Self as Opposed to Agent-Self

Even if there aren’t agent-selves in Spinozism, there can still be an encompassing subjective center—a perspective—to a human mind which can be called a witness-self.

Let us return to the tentative understanding of selfhood. For many, and certainly for Koistinen (2009), a self is what explains why some collection of thoughts or acts belong to one individual. The self is as it were is the central point around which these acts gravitate. But what exactly does it take for there to be such a gravitational centre for some set of acts? If one explains selfhood in terms of agency, then this means for the self to be the source of the actions deliberated and carried out by the individual in question (Agent Causation). What criteria should the self, qua source of its actions, meet? On the very least, I believe, such an agent-self should not itself be an act. Yet, Spinoza refers human minds to the same ontological level as their ideas.21 For this reason, human beings are not, in the final analysis, the thinkers of their thoughts or the agents of their acts. They cannot unify their acts by being the ones ultimately acting them out.

If, however, the self is instead more intimately connected with witnessing phenomena than with carrying out acts, we can define selfhood alternatively in terms of a subjective center or perspective, but not in terms of agency. With this difference in mind, my view can be articulated as follows: we can have adequate ideas of how our acts do not ultimately belong to us. These ideas are adequate in the sense that they can be understood through our essence exactly since they posit our essence as intimately dependent upon God: indeed, as being, somehow, God. Such adequate ideas allow us to witness the evanescence of falsely posited agent-selves. We become ourselves (witness-selves) in adequately thinking that we are not ultimately the thinkers of our thoughts. This view—that we are not fundamentally agent-selves and that we become ourselves (our witness-selves) most fully when we acknowledge this fact—may seem controversial. In the remaining section, I will defend it by showing how Spinoza draws several significant ethical and soteriological benefits from it in the Second Part of his Ethics.

It should be pointed out that this reading—that if there is human selfhood it should consist primarily in witnessing being acted out or in intellectually relinquishing one’s agency to God—may appear incompatible with Spinoza’s contention that human beings can be the adequate causes of at least some effects (3d2). In this regard, it should be noted that to explain how human beings can ultimately be adequate causes is a general problem arising for any coherent interpretation of Spinozism; it is not engendered by this specific proposal.22 Nonetheless, I do believe that my reading is not entirely helpless in responding to the challenge. For I have not denied that human beings can act insofar as they are God somehow: what I have rejected is more precisely that that intersection between God and human where action takes place can be described as a human self.

4. On the Value of Being an Act Rather than an Agent

The view I advance does not prevent us from having adequate ideas about who we fundamentally are, since we can have adequate ideas that originate in what we fundamentally are when we realize that insofar as we act, we are not fundamentally the independently acting agents we perceive ourselves to be, but rather God. This position may seem outrageous. It is therefore important to note that Spinoza considers the insight in question to be of ethical and therapeutical value.

In 2p49, Spinoza asserts that ‘In the Mind, there is no volition, or affirmation and negation, except that which the idea involves insofar as it is an idea’. As pointed out previously, this proposition in and of itself makes it implausible that selves should be mental agents in Spinozism, because what job is there for a self to do that is not already accomplished by the ideas constituting the mind for which the self is posited? For Spinoza, our inner life is constituted by ideas. Ideas are not the objects of thinking, they are instances of the activity of thought. God, and not the self, is the ultimate agent of these acts. One may perhaps find such a reductionist view dissatisfying from an ethical perspective. It is therefore important to bear in mind that Spinoza considered it empowering.

More precisely, Spinoza ends the long scholium to the cited proposition by commenting on ‘how much knowledge of this doctrine is to our advantage in life’ (2p49s[IV]). All four advantages Spinoza lists are practical. More precisely, the first advantage he lists is mystical or philosophical, the second emotional or therapeutical and the additional two each have bearing on our social being and political philosophy. For reasons of scope, I will only comment on the former two here.

First, the doctrine that we do not think our ideas but our ideas are thought by God is to our advantage in life ‘[i]nsofar as it teaches that we act only from God’s command, that we share in God’s nature, and that we do this the more, the more perfect our actions are, and the more and more we understand God’ (2p49s[A]).

2p49 is not only the contention that ideas are volitional in addition to being ‘ideational’. It is also, and more importantly, the contention that ideas are themselves acts—slices of the activity of thinking, as it were—and that they are thus not the objects of thinking. They are acts or modes of God: ways in which God causes himself to be. It is on account of this that they are necessitated by God or as Spinoza puts it, ‘at God’s command’. The reason why everything is at ‘God’s command’ is not merely ‘mechanistic’. That is, everything is not ‘at God’s command’ because God is the prime mover, initiating a deterministically operating series of events. (For the contention that there is no prime mover, see 1p28.) Instead, the reason why everything is ‘at God’s command’ is ontological. We can ‘share in God’s nature’ only because we are, as it were, the perfective aspect of his acting: the acts his infinite activity results in. And since his activity is necessary, so are we, the acts he performs. In saying that ‘we act only from God’s command’ since we, insofar as we act, are God, Spinoza seeks to redeem our agency. But in so doing, something is also irretrievably lost: the agent-self.

If we do not posit selves as agents, it appears that we cannot hold each other truly responsible for what we do or think. This is an enormous cost for an ethical system. But 2p49 is designed to eliminate one form of responsibility, namely the responsibility that can arise in the space between the idea and the subject affirming or negating it. Since ideas are as it were automatically self-affirming and self-denying, there is no need for a subject to do the job the ideas themselves already perform. Spinoza does not deny that there is phenomenological personhood: for there is self-awareness to each set of ideas that constitutes a mind. What he denies is that, ultimately, a human self does the thinking. The mind is instead ontologically on a par with the ideas that constitute it. As a consequence, we lose the mastery we thought we had over our ideas (and other actions). But by realizing that we are God’s actions and that all our activity is ultimately predicated of him we can come to consciously ‘share in’ his being. Spinoza, on this reading, offers the sense of mastery we had over our thoughts and actions on the altar of sublime unity. We are not thinkers: we are thoughts God thinks. But since we are self-aware thoughts, it is possible for us to realize this and consciously enjoy being acted out by God. It is exactly in realizing that we are self-aware thoughts, that we can enjoy freedom: we become free in realizing the full extent of our ‘slavery’, by which I mean our utter dependence on God.

Second, Spinoza says that the doctrine is of advantage to us in life

since it teaches us how we must bear ourselves concerning matters of fortune, or things which are not in our power, i.e., concerning things which do not follow from our nature—that we must expect and bear calmly both good fortune and bad. For all things follow God’s eternal decree with the same necessity as from the essence of a triangle it follows that its three angles are equal to two right angles. (2p49s[B])

Spinoza has established—or at any rate thinks that he has established—that all things are necessary already in 1p29. So why does that doctrine—the necessity of things—reappear at this point in the exposition? On my reading, it does so because by 2p49 Spinoza believes that he has shown that not only do all things operate by divine necessity: but that so does the concatenation of ideas in our minds. By realizing this—that is, by the occurrence in the self-cognizing idea of the set that constitutes our mind of an idea that realizes this—Spinoza thinks that we can develop a Stoic calm in the face of external events as well as in the face of our own feelings and thoughts. By relinquishing responsibility for our inner lives—by acquiescing in the fact that we are not the agents of our thoughts but merely the ones witnessing them—we become more at peace.

Interestingly, Spinoza here enjoins upon us a love of necessity that is as it were infinitely folded. We are to acquiesce in the fact that our inner lives are necessitated. We are to acquiesce in the fact that this act of acquiescence too was necessitated, and so on, ad infinitum.23 Perhaps there would be no practical significance to our coming to accept that we necessarily think our thoughts—that, indeed even our acceptation of this fact is necessitated—if the necessity by which we think was just natural. But the reason why things are necessitated is divine as well as natural. Hence, Spinoza hypothesizes that upon realizing it, we will reach the calm of the person who neither hopes nor fears, but simply loves reality, since reality is God.

5. Final Remarks

Others have done an excellent job in exploring the continuity of the early modern ‘invention’ of autonomous selves in our contemporary intuitions. In this essay my goal was not to demonstrate the continuity between contemporary Western sensibilities and early modern philosophy, but instead to point out a potential discontinuity. This discontinuity is potentially of comparative interest, for it is conceptually kindred to Buddhist views on the self, or rather, the non-self (anatta).24 I reserve a comparative investigation for a future paper.

Independently of any comparative value my essay may have, its major thesis is of interest to anyone who suspects that the self is a purely theoretical entity. For Spinoza, there is, admittedly, a phenomenological person we encounter in the distorted mirror of introspection. But that person does not need a self to subsist. At the most fundamental level, we are not ourselves: we are not the agents or thinkers that we perceive ourselves to be. Although this means that there is no room for agent-selfhood in Spinoza, another kind of selfhood is still within reach.

On the view I propose, one way for human beings to have adequate ideas is by witnessing how God acts through them. In so doing, they do not ‘appropriate’ divine action but acknowledge that insofar as they act, they are not the agents they perceive themselves to be, but rather God. Of course, this view is prima facie paradoxical: I claim that human beings can have adequate ideas and thus act exactly in acknowledging that their action is not theirs. Yet, the prima facie paradoxical view I defend is compatible with Koistinen’s important insight that what we fundamentally are, is God (modified somehow). What I dispute is that, on such a scenario, we are fundamentally the selves we think ourselves to be: the agents of our actions and the thinkers of our thoughts. In the final analysis, we are instead acts performed by a divine agent. In conclusion I submit that, for Spinoza, for so long as we think that we are our (agent-)selves, our understanding of what we fundamentally are is inadequate. We become our (witness-)selves in witnessing how our agency fades into God’s agency.


  1. For example, a widespread assumption in contemporary practical philosophy is that people are autonomous moral agents, that is, persons whose reasoning is in principle sensitive to moral reasons. According to Schneewind (1998) this was not always so: the conception of morality as autonomy (rather than, say, obedience) was engendered with Kant. [^]
  2. I adopt the shorthand where 1p14 refers to the 14th proposition of the first part. app = appendix, a = axiom, c = corollary, d = demonstration (when following a proposition-indicating number) or definition (when following a Book-indicating number), l = lemma, p = proposition, s = scholium. For the English translation of Spinoza’s texts, I refer to Edwin Curley editions of 1988 and 2016, henceforward C. I and II. For the Latin original, I consult Gebhard’s edition from 1925. [^]
  3. Koistinen cites Della Rocca (1996: 41–43) as a proponent of the view. This appears indeed to be the view that first-time readers of Spinoza are served. See for instance the introduction of individuality proposed by Beth Lord (2010). For a more explicit proponent of the bundle theory, see Williams (2010). [^]
  4. A similar interpretation is defended by Sangiacomo and Nachtomy (2018). [^]
  5. Consider for example Peter van Inwagen’s account of agent causation. It is the kind of causation that is rooted in selves. Selves are posited exactly to allow for agents to freely cause effects. [^]
  6. In 2p15 Spinoza writes: ‘The idea that constitutes the formal being [esse] of the human Mind is not simple, but composed of a great many ideas.’ One may perhaps want to object by saying that one cannot infer that the mind itself is complex from the fact that the-idea-that-constitutes-the-formal-being-of-the-mind (whatever this is) is complex. Yet, within the confines of Spinozism, this inference is possible. For the mind is an idea, and as an idea, it is numerically identical with the idea of this idea (2p21s). But the idea of the idea is nothing but the formal being of the idea, that is, the idea of which Spinoza speaks in 2p15. Yet, one puzzle remains: why does Spinoza speak at all about ideas of ideas, if these for all purposes are identical with the ideas they intend? [^]
  7. For a recent interpretation which construes Spinoza as a structuralist vis-à-vis essence, see Morrison (2018). [^]
  8. For such a view, see Fodor (1981: 22). [^]
  9. Leibniz was famously suspicious of this claim. For Leibniz, ‘Ideae non agunt Mens agit’. See the ‘useful caricature’ provided by Laerke (2008: 38–39). One should be careful before one adopts the Leibnizian position as the one dictated by ‘common sense’ (as Koistinen does in suggesting that as the bundle theory cannot explain why a mind should belong to a subject, one should ‘avoid attributing such a view to any great philosopher of the past’ [2009: 151]). For it is at least possible that our synthetic consciousness (our experience of ourselves as subjects and agents) is epiphenomenal and secondary to (supervenient upon) neurological phenomena of a different kind, such that it is really the ‘ideas’ that think, although the ‘mind’ perceives itself as the one thinking them. [^]
  10. For consider that, for any belief, we can test whether its truth appears to us. If it doesn’t, it’s false. But we don’t have the means to verify our beliefs in this way. [^]
  11. Similarly, Matheron (1997: 233) thinks that Spinoza is hereby compelled to the existence of a human mental agent distinct from its ideas. [^]
  12. An Essay Concerning Human Understanding, II.27.9 (Locke 1959). [^]
  13. It has been objected that the memory relation cannot account for personal identity, and that other relations have been proposed as criterial in its place, such as for instance psychological connectedness (Parfit 1984: 205–6). [^]
  14. Koistinen subsumes memory under the imagination, but clearly one remembers what one considers past. Yet, for Spinoza, it seems that the imagination is the activity of perceiving things as present (2p17cds). So perhaps one would be inclined to dismiss the subsumption of memory under the imagination. However, presence before the mind need not involve the perception of simultaneity; on Spinoza’s view, a thing is perceived as present by a mind that does not reject the existence of the thing in question. So, for instance, things we fantasize are ‘present’ before our minds. More precisely, to imagine is exactly to ‘[regard] external bodies through ideas of the affections of its own Body’ (2p26cd). What the memory associates is the presences of things perceived through the affections the body undergoes, and ideas of such affections are called images (see 2p18s). It is thus exactly because images are impressed on the mind and not actively constructed by it that memory is passive. More precisely, Spinoza defines memory (in 2p18s), as ‘a connection that is in the Mind according to the order and connection of the affections of the human Body’, and images are produced by the same (passive) affections. Therefore, this connection between ideas is to be distinguished from the order that the intellect actively establishes by probing the natures of things (loc. cit.). [^]
  15. On the very least, this appears to have been Spinoza’s own view on personal continuity, as expressed in his discussion of the Spanish poet (4p39s). [^]
  16. In the Ethics, the attributes are identified with God’s essence a sufficient number of times that one is justified in assuming that they are indeed identical with God’s essence. It follows that the relation whereby God is identical with his essence is not transitive. Bledin and Melamed (2019) have addressed this infamous problem by proposing a tripartite identity relation in Spinozism. [^]
  17. Viljanen (2007) provides a full account for how action and essence are related in Spinozism. [^]
  18. The following important objection should be addressed. There appears to be, for Spinoza, a necessary feature of the mind, a core of the mind, as it were, on whose continued existence the continued existence of the mind depends. This is the formal essence of the mind, the idea of the mind, a notion with which Spinoza’s doctrine of eternity is intimately linked. In response to the idea that the formal essence of the mind constitutes a necessary ingredient or a ‘core element’ of the mind, I would like to object that the mind cannot be grounded in the idea of the mind. This cannot be the case if we consider the reality of representation to be such that it is at least partially grounded in the reality of what is represented—that is, the grounding of the identity of the mind in the idea of the mind appears untenable from a realist position. Nor does it appear to be tenable within the framework of Spinoza’s metaphysics. For, if I interpret his second parallelism correctly, the idea of the idea is identical with the idea of which it is the idea. That the continued existence of x is necessary for the continued existence of x is nothing but a tautology and does not show that for Spinoza, there is a ‘necessary ingredient’ or a core element to the mind over and above the mind itself. I acknowledge that I have not been able to reconcile Spinoza’s second parallelism with the interpretation that the formal essence of the mind is also the part of the mind that is eternal. [^]
  19. I thank an anonymous referee for pressing this issue. [^]
  20. Koistinen’s idea merits being cited in full. It reads: Without going into the details of this suggestion, I just want to point out that for Spinoza all ideas belong to natura naturata, and not to natura naturans, which means that they are caused by God and hence conceived through God; this also means that the adequate ideas of human beings should be conceived through God’s nature, which suggests that human beings cannot act in having adequate ideas if human beings do not have an idea-free nature that somehow is identical with God’s nature. Thus, our nature, which is causally responsible for the fact that we act, cannot be constituted by ideas of any sort, and our subjecthood cannot be reduced to ideas or to bundles of them. (Koistinen 2009: 153) [^]
  21. For a historical version of the view that agent and act don’t belong to the same ontological category, consider the Aristotelian dictum that actus sunt suppositorum—acts belong to supposita (endorsed by, for example, Aquinas [1980b]: ST I q. 39 art. 5 ad. 1). A suppositum is the entity which underlies the action or quality expressed by the predicate. Related passages in the works of Thomas Aquinas indicate which kind of things can serve as supposita. Supposita are singular subsisting things (actus sint singularium subsistentium: ST I, q. 56, art. 1. ob 1) or individuals (actus individuorum sunt: In II. Sent. ds. 32, q. 1, art 2, c [1980a]). In brief, supposita are substances: things that subsist on their own. Agents are the supposita of predicates expressing actions. Hence, agents are substances. [^]
  22. See, for instance, Marshall (2013: 22–23), who articulates the problem in these terms: ‘(a) Having an adequate idea of x requires having ideas of all x’s causal antecedents; (b) Every finite mode has an infinite chain of causal antecedents; (c) No human mind can have ideas of an infinite chain of causal antecedents. […]; (d) Therefore, no human mind can have an adequate idea.’ See also Della Rocca (1996: 183, n. 29) and Steinberg (1981). Both Steinberg (1981) and Marshall (2013) argue that adequate knowledge is possible, since it primarily targets of formal essences. Such knowledge need not involve infinite causal regresses. [^]
  23. The reader may here recognize the ‘infinite loop of eternal return’ to which Löwith (1997) devoted his study of Nietzsche. [^]
  24. With the important exception of the so-called Puggalavāda schools. I do not claim to expound Buddhist doctrine in any way in this essay. For an accessible study see Collins (1992). [^]


I would like to acknowledge The Hong Kong PhD Fellowship for funding.

Competing Interests

The author has no competing interests to declare.


1 Aquinas, Thomas. 1980a. In Quattuor Libros Sententiarum. Vol. 1 of Thomae Aquinatis Opera Omnia, edited by Roberto Busa S.I. Stuttgart-Bad Canstatt: Frommann-Holzboog.

2 Aquinas, Thomas. 1980b. Summa Contra Gentiles; Summa Theologiae. Vol. 2 of S. Thomae Aquinatis Opera Omnia, edited by Roberto Busa S.I. Stuttgart-Bad Canstatt: Frommann-Holzboog.

3 Bledin, Justin and Melamed, Yitzhak Y. 2019. “Ethica more logico demonstrata” (manuscript).

4 Collins, Steven. 1992. Selfless Persons: Imagery and Thought in Theravada Buddhism. Cambridge: Cambridge University Press.

5 Della Rocca, Michael. 1996. Representation and the Mind Body Problem. New York: Oxford University Press.

6 Fodor, Jerry. 1981. Representations. Brighton: Harvester Press.

7 Koistinen, Olli. 2009. “Spinoza’s Eternal Self.” In Topics in Early Modern Philosophy of Mind, edited by Jon Miller. Vol. 9 of Studies in the History of Philosophy of Mind. Dordrecht: Springer. DOI:

8 Laerke, Mogens. 2008. Leibniz lecteur de Spinoza: La genèse d’une opposition complexe. Paris: Honré Champion.

9 Locke, John. 1959. An Essay Concerning Human Understanding. New York: Dover Publications.

10 Lord, Beth. 2010. Spinoza’s Ethics. Bloomington: Indiana University Press.

11 Löwith, Karl. 1997. Nietzsche’s Philosophy of the Eternal Recurrence of the Same. Translated by Harvey Lomax. Berkeley: University of California Press. DOI:

12 Marshall, Eugene. 2013. The Spiritual Automaton. Oxford: Oxford University Press.

13 Matheron, Alexandre. 1997. “L’amour intellectuel de Dieu, partie éternelle de l’‘amor erga deum’.” Les études philosophiques 2: 231–48.

14 Melamed, Yitzhak. 2013. Spinoza’s Metaphysics: Substance and Thought. New York: Oxford University Press. DOI:

15 Morrison, John. 2018. “Two puzzles about Thought and Identity in Spinoza.” In Melamed (ed.) Cambridge Critical Guide to Spinoza’s Ethics. Cambridge: Cambridge University Press.

16 Parfit, Derek. 1984. Reasons and Persons. Oxford: Oxford University Press.

17 Sangiacomo, Andrea and Ohad Nachtomy. 2018. “Spinoza’s Rethinking of Activity: From the Short Treatise to the Ethics.” The Southern Journal of Philosophy 56 (1): 101–26. DOI:

18 Schneewind, Jerome B. 1998. The Invention of Autonomy. Cambridge: Cambridge University Press.

19 Spinoza, Baruch de. 1925. Opera. 4 Vols. Edited by Carl Gebhardt. Heidelberg: Carl Winter.

20 Spinoza, Baruch de. 1988–2016. The Collected Works of Spinoza. 2 Vols. Edited and translated by Edwin Curley. Princeton: Princeton University Press. DOI:

21 Steinberg, Diane. 1981. Spinoza’s Theory of the Eternity of the Mind. Canadian Journal of Philosophy 11 (1): 35—68. DOI:

22 Viljanen, Valtteri (2007). Field Metaphysic, Power, and Individuation in Spinoza. Canadian Journal of Philosophy 37 (3): 393–418. DOI:

23 Williams, Caroline. 2010. “Affective Processes without a Subject: Rethinking the Relation between Subjectivity and Affect with Spinoza.” Subjectivity 3 (3): 245–62. DOI: