Support 100 years of independent journalism.

26 February 2020

The paradox of an atheist soul

Why the idea of a single self only makes sense in a theistic world. 

By John Gray

There are many arguments for theism, most of them not worth rehearsing. The ontological argument, first formulated by St Anselm in the 11th century and reframed by the 17th-century French rationalist René Descartes (1596-1650), maintains that God must exist because humans have an idea of a perfect being and existence is necessary to perfection. Since many of us have no such idea, it is a feeble gambit. The arguments of creationists are feebler, since they involve concocting a theory of intelligent design to fill gaps in science that the growth of knowledge may one day close. The idea of God is not a pseudo-scientific speculation.

A different and more interesting approach is to argue that theism is suggested by the fact that we experience ourselves as unified, conscious beings – in other words, as having a soul. Not necessarily an immaterial entity, the soul is the part of us that strives to realise what is best in our nature. We do not come to know the soul through any special revelation. We know it by considering the kind of creature we find ourselves to be – a thinking being inhabiting a life-world that seems to reflect a mind greater than our own. Once we realise we have a soul, theism becomes a credible way of thinking.

Such is the approach adopted in this lucid and illuminating book by John Cottingham, professor of the philosophy of religion at University of Roehampton. Modestly described as an essay, Cottingham’s short study explores fundamental questions more fully than many much longer volumes. While it fails as an argument for theism, it is forceful and compelling in arguing that the idea of selfhood taken for granted in secular societies makes sense only in the context of a theistic world-view.

Cottingham presents a version of the transcendental argument deployed by the German Enlightenment philosopher Immanuel Kant (1724-1804). A transcendental argument does not appeal to anything factual. Instead, it asks what must be true if certain features of human experience are accepted as given. Kant used it to support his belief in a universal moral law and, at points in his writings, the existence of God. As used by Cottingham, its purpose is to refute the Scottish sceptic David  Hume (1711-1776), whom Kant described as “having interrupted my dogmatic slumber”. In A Treatise of Human Nature (1739), Hume had written that the self is “nothing but a bundle or collection of different perceptions, which succeed one another with an inconceivable rapidity, and are in a perpetual flux and movement”. If the self is not an autonomous entity but an assemblage of sensations Kant’s theistic faith crumbles into dust.

Cottingham spells out the connection between theism and the idea of the self:

Sign up for The New Statesman’s newsletters Tick the boxes of the newsletters you would like to receive.
I consent to New Statesman Media Group collecting my details provided via this form in accordance with the Privacy Policy

It is a fundamental theistic belief, following the words of Genesis, that human beings are made “in the image” of God; and this is taken to be especially true in virtue of our conscious minds, in virtue of our attributes of intellect and will. Theism thus posits a source of ground of all being that is somehow mind-like: consciousness is taken to be at the heart of reality. The theistic picture tends to be discarded or ignored by the majority of contemporary philosophers, but it seems perverse to dismiss it from consideration should it turn out to fit rather well with certain aspects of reality that cannot in integrity be denied… [such as] the irreducible reality of consciousness.

Theistic religions are inherently anthropocentric. God is an infinitely enlarged projection of human personality. Yet many religions have understood God as an impersonal world-soul that may spawn souls that resemble human beings, but is itself remote from anything human. Other religions have done without any soul at all.

Older than Christianity and at least as philosophically sophisticated, Buddhism begins by rejecting the concept of the soul. The core Buddhist doctrine of anatta (no-self, no-soul) teaches that there is nothing in humans like a continuing identity. Popular Buddhism upheld older ideas of metempsychosis, according to which a soul is reincarnated after death. But in Buddhist philosophy, only a complex psychophysical process continues from death. Wherever it seems to exist, selfhood is an illusion.

The Buddhist view is similar to Hume’s: the apparent solidity of the self comes from the extreme rapidity with which one perception follows another. One of the goals of Buddhist meditation is to slow this process, so the practitioner can shed the illusion of selfhood. Some Western mystics have talked of the individual soul merging with a world-soul, but in Buddhism the idea of a world-soul is also rejected. Human salvation involves ridding oneself of any idea of the soul, human or divine. It is hard to think of a view more distant from the central traditions of theism.

When exploring the idea of the soul Cottingham says nothing of Buddhism, or any non-Western religion. He considers briefly a modern version of the denial of self- hood, which questions the idea that we should aim for narrative unity in our lives. Any such defence of the “episodic” or  “happy-go-lucky” life, he tells us, “seems open to a swift and devastating rebuttal: lives of this episodic kind are possible only because others who are not leading happy-go-lucky lives are sustaining the stable relationships that make their easy-come-easy-go attitude possible”. He goes on to observe that advocates of the “episodic” life “tend to be drawn in the end to abandon the very idea of a self persisting over time… Yet the more we think about this, the more it starts to look like a fantasy of evasion.”

Cottingham’s rebuttal may be swift but it is hardly devastating. An episodic view of human life is the claim that no persisting self is revealed in the course of our lives. Certainly there are patterns of continuity  in memory and behaviour, but these marks of selfhood shift and fade in lives that are long or varied.

The life of Bertrand Russell (1872-1970) was both. He writes in his autobiography that when he looked back he found not a single person but something more like a club whose members changed over time. The solitary, rationalistic and rather puritanical self of Russell’s late Victorian youth was not the self that flirted with mysticism as he fell unhappily out of love with his first wife. Nor was it the self that emerged from a spell in prison for pacifist resistance against the First World War, after which his interests shifted from mathematics and logic to politics, and he travelled to Lenin’s Russia and war-torn China. Still less was it the self that married three more times and had countless affairs. Reflecting on his life, Russell found no enduring selfhood.

Of course, Russell was exceptional in many ways. But an episodic life featuring a succession of disparate selves captures the experience of many people better than any story of the continuous unfolding of an autonomous individual. The selfhood that some find throughout their lives is a by-product of stability in society, which rarely lasts for very long. War, revolution and social breakdown regularly overwhelm the sense of being a person with a coherent life-story. A unitary self is a fantasy that can be enjoyed only in peaceful times.


Cottingham’s dismissal of the episodic self is of a piece with his theory of value. He points out that morality makes demands on us we are compelled to recognise, whether we like them or not. Accordingly, human values reflect “an objective, authoritative demand” imposed on us by what he describes as “strong normativity”. No doubt many people experience morality in this way, but that does not mean values are objective. Their apparent objectivity is a projection from the ways of life by which human beings are formed. There are many moralities, each experienced as compelling by their practitioners.

Even within the Western tradition, as Tom Holland showed in Dominion: the Making of the Western Mind, there are enormous moral gulfs. The Iliad knows nothing of forgiveness, nor does Aristotle’s Ethics of humility. Self-sacrifice figures nowhere in the Epicurean pursuit of tranquil pleasure, nor does concern for the downtrodden and forgotten in Stoicism. Our revulsion at the gladiatorial games of ancient Rome does not come from any inbuilt repugnance at the spectacle of human suffering and violent death. There is no sign that those who watched the games felt any such revulsion. Nor is there much evidence from that era that slavery was felt to be inherently wrong. The repugnance we feel for these practices is an inheritance from Jewish and Christian ideas of human dignity and equality.

In this and other cases, what liberal humanists believe to be universal values are relics of particular religious traditions. Here Nietzsche was right. Human values are too changeable, and too divergent, for morality to be in any meaningful sense objective.

Philosophy did not cease to be a handmaiden of theology when it abandoned  theism. In his monumental study On  What Matters (2011), the late Oxford philosopher Derek Parfit tried to show that some things really, truly, deeply matter; that there are moral truths that give us reason to act and live in certain ways. But in what kind of world could such truths exist and possess authority over human beings? Without theism, or some Platonic spiritual realm, these supposed objective values are left hanging in the void.

Human values need not be exclusively human. As Darwin implied in The Expression of The Emotions in Man and Animals (1872), our moral responses can be traced back to our evolutionary kin. Without having the kind of self-conscious awareness humans intermittently display, other animals inhabit life-worlds and pursue distinctive ideas of good. Cottingham comes close to denying this evident fact, speculating that without human beings the world would be “blank, silent, dark”. He writes us that “no animal could formulate the Cartesian idea of ‘this me by which I am what I am’”. (It is unclear what this is meant to prove. Possibly it shows only that other animals are not as prone to delusion as human beings.) But there is nothing to suggest values subsist in a realm beyond that of living organisms.

Secular thinkers who cling to the idea of human autonomy have not shaken off theism. Cottingham writes, “The contrast  between the theistically inspired and the post-Enlightenment conceptions of the role of the self could not be more marked.” Actually, the opposite is the case. As Cottingham acknowledges a page later, it was Kant – a lifelong Christian – who asserted the prototypical Enlightenment belief in “independent human rationality and autonomy”. The belief that human beings are essentially autonomous agents is the theistic myth of the soul reiterated in rationalist terms.

When philosophers deploy transcendental arguments it is in order deduce themselves as they imagine themselves to be. Because Cottingham argues that the experience of selfhood points in the direction of theism, secular readers may decide that  his essay is not for them. But if they do they will be mistaken, for the sense of self he invokes is their own. 

In Search of the Soul: A Philosophical Essay 
John Cottingham
Princeton University Press, 192pp, £18.99

Topics in this article:

This article appears in the 26 Feb 2020 issue of the New Statesman, The death of privacy