When we are confronted with a question to which we want an exact, or at least profound, answer, and the question is posed in natural language, a useful first step is to remove as much ambiguity as possible. In logic we do this by formalising the question: deciding what its terms actually mean and what would count as an answer.
Consider the notoriously vague but apparently profound question: What is the meaning of life?
Two words immediately cause trouble: meaning and life.
Start with meaning. The English word itself has several quite different senses. The most straightforward concerns reference or denotation. Suppose somebody encountering Hegel meets the German word Seele but does not speak German. They ask: “What does Seele mean?” The rough answer is soul.
We give this a simplified model-theoretic description. There is an expression in one language whose interpretation the questioner does not know. There is some domain of things or concepts about which the language speaks, together with an interpretation function connecting linguistic expressions with that domain. The English speaker already possesses the corresponding machinery for English. We solve the problem by supplying an English word, soul, whose semantic interpretation sufficiently overlaps that of Seele.
Natural language is much messier than Tarskian semantics, of course. Translation involves history, context and connotation as well as denotation. But the logical picture will do: an unfamiliar expression is connected, through an interpretation, with something already understood.
This is plainly not what somebody means when asking about the meaning of life. They are not asking what object the word life denotes.
There is another sense of meaning: significance or intent. The classic joke concerns Talleyrand, the notoriously subtle French statesman. On hearing that Talleyrand had died, Metternich is supposed to have remarked: “I wonder what he meant by that.”
Metternich knows perfectly well what death is. His supposed problem is to discover how this particular event fits into some larger pattern of intentions, plans and consequences.
That gives us a different structure. We have an event or action embedded in a larger setting, together with an implicit theory of that setting. People perceive things, acquire beliefs, formulate intentions, pursue goals and act. We understand an action when we can locate it within this wider scheme. Somebody abruptly leaves a meeting; once we learn that their house is on fire, the behaviour acquires its significance.
Meaning in this sense is relational. Something has meaning because of the place it occupies within some wider explanatory structure.
That is much closer to what people are asking when they enquire after the meaning of life. They want to know what life signifies: what larger structure it belongs to, what role it plays within that structure and, perhaps, towards what end the whole arrangement is directed.
Unfortunately this merely transfers the difficulty. With Talleyrand we already have the larger political framework. With life as a whole, it is precisely the larger framework that is missing.
Then there is the second troublesome word: life.
If by life we mean simply what human beings have in common with tadpoles and cabbages, the question has several answers, none of them particularly interesting. One naturalistic account begins with physics.
Given a universe initially very far from thermodynamic equilibrium, continuing flows of usable free energy, suitable chemistry and the right local conditions, highly organised structures can arise. Once some of them acquire replication, heredity and variation, Darwinian evolution takes over.
On this account life is one of the interesting things matter does under particular conditions. It is elaborate non-equilibrium physics which, once replication appears, bootstraps itself through natural selection.
This may be broadly right as an account of biological life. It is not usually taken to be what the questioner really wanted.
The intended sense of life is normally much narrower and more peculiarly human. It concerns the existence of a conscious social being possessing a conception of themself, giving an account of themself to other people, making choices and seeing those choices as forming some kind of trajectory.
We speak of a family life, a career, a vocation, achievement, failure, loyalty, betrayal and purpose. A human life is not construed as metabolism continuing for eighty years. It's a project.
We can therefore reformulate the original question:
What is the significance of the life of a human being reflecting upon their situation and available choices?
It also contains an assumption about value. The person asking does not normally expect to be told that every possible life is equivalent. Become a surgeon, raise children, collect stamps, spend forty years intoxicated, torture strangers, kill yourself - it makes no difference?
We do not experience our possible futures as occupying a flat value space. Some possibilities appear worthwhile, others contemptible; some fulfilling, others disastrous. Asking for the meaning of life therefore amounts to asking for some framework within which these differences in value can be understood.
Religion supplies one candidate framework. Human existence forms part of a created order; that order has a purpose; human beings have a nature and an intended end; certain ways of living conform to that order better than others and are virtuous.
In the terminology of the present argument, religion supplies the missing boundary conditions.
Remove it and matters become less comfortable.
Another answer is nihilism. Conclude that there is no encompassing structure. Human beings manufacture local purposes, but the universe supplies none. Asking for the meaning of life is therefore rather like asking for the foreign policy of a cabbage: the question presupposes a level of reality-embedding which does not actually exist for the vegetable.
There is, however, a less drastic naturalistic answer. We can relativise our values to the kind of creatures we actually are.
Human beings are evolved organisms carrying a characteristic collection of needs, emotions and motivations: attachment, sexuality, parental concern, reciprocity, curiosity, status, play, achievement, fear, anger, compassion and so forth.
These are not at all decorative additions to an otherwise neutral reasoning machine. They are the primary machinery by which anything matters to us in the first place.
Evolution explains why this machinery has approximately the shape it does. Organisms wholly indifferent to survival, danger, offspring and cooperation have an obvious tendency not to leave descendants. Their motivational architecture removes itself from the population.
This does not mean that whatever evolution produced is morally right. Natural selection is not a moral philosopher. The more modest point is that any account of a satisfactory human life must begin with the creature whose life it is. A viable conception of human flourishing cannot be radically detached from human biology, psychology and social nature.
This gives us a naturalistic answer to the question, but only a relative one.
These things are not important because the universe regards them as important. They are important to creatures like us because of what creatures like us are.
There is nevertheless one harder constraint.
Existence is the precondition for every other project, value and experience. For any system of values to remain instantiated in the universe over evolutionary time, some continuing lineage of agents capable of holding those values must persist. A value system which systematically causes its bearers to disappear does not thereby become false; it simply ceases eventually to have any bearers.
That sounds almost trivial. It becomes less trivial once intelligence acquires the ability to alter its own motivational machinery.
Human nature is not logically necessary. It is the contingent product of a particular evolutionary history of the species. Advanced genetic engineering, neurotechnology and artificial intelligence may eventually allow us to alter the very neural mechanisms upon which our values subsist.
At present there is a rough coupling between affect and biological success. Sex feels good; attachment feels important; children provoke extraordinary investment; achievement, status, curiosity and social belonging motivate enormous effort.
Evolution did not give us these experiences in order to make us happy. It produced organisms in which such proximal motivations tended, statistically and imperfectly, to generate behaviour compatible with survival and reproduction.
An intelligent species can eventually do something about this trick.
Why undertake the cumbersome business of strenuous achievement in order to experience satisfaction if satisfaction itself can be engineered?
Why pursue the complex business of sexual relationships if the associated affect can be generated directly?
Why endure the expense and inconvenience of children if parental fulfilment can be perfectly simulated?
Once the proximal psychological reward can be detached from the behaviour for which evolution originally selected it, a dangerous possibility appears. The organism can consume the reward while abandoning the very activity which made the reward system evolutionarily viable in the first place.
Imagine beings engineered to experience intense positive affect directly and indefinitely. They might inhabit perfect virtual worlds, experience astonishing happiness, reproduce rarely or not at all. Eventually all such beings disappear.
From inside their altered value system there need be nothing wrong with this.
This is not peculiar to humanity. It is a general problem for any intelligent lineage produced by evolution.
Natural selection builds agents whose immediate motivations are proxies for outcomes which tended to generate reproducing descendants. But sufficient intelligence and technology eventually allow those agents to understand and manipulate the proxies themselves.
At that point evolution has produced something capable of defeating the mechanism by which evolution produced it.
There is no logical reason why this must lead to extinction. A civilisation might deliberately preserve reproductive and expansionist drives, or redesign itself with long-term persistence as an explicit goal. But the coupling is no longer automatic. For the first time, perhaps, the continuation of the lineage becomes an engineering decision. The abyss remains.
In the very long run, the universe will belong not necessarily to the creatures with the truest values, but to those whose values permit them to remain in it.
P(doom): Where Are They?
There is an obvious speculative connection here with the Fermi paradox.
A sufficiently advanced civilisation may eventually acquire the ability to redesign not merely its environment but its own affective machinery. Indeed it would be surprising if it did not. The reward systems left to us by evolution are makeshift, frequently unpleasant and spectacularly inefficient and hard work. Any technologically mature species might reasonably wish to improve them.
But that creates the possibility of a peculiar Great Filter. Civilisations may discover how to optimise subjective experience before they have secured indefinite survival. Instead of expanding through the galaxy they turn inward, uncouple reward from reproduction and exploration, and disappear in perfect contentment.
This cannot by itself be a complete solution to the Fermi paradox. It would have to happen with an extraordinarily high probability. Even one civilisation retaining a strong drive towards expansion could eventually become conspicuous on astronomical scales.
But it reminds us of the general principle. Intelligence does not merely increase a species' power to survive. Eventually it gives the species power over the very desires which caused its ancestors to survive in the first place. Use with care.
