The evolutionary origins of subjectivity and agency

The concept of individuation is at the base of a conceptual structure that runs from agency through autonomy and liberty to the idea of the state, law and political legitimacy.

These concepts belong to the Geisteswissenschaften, (there isn’t a properly equivalent term in English), the intellectual disciplines concerned with intellectual and cultural entities and events, rather than the Naturwissenschaften, the intellectual disciplines, the natural sciences, that are concerned with natural entities and events.

However, the form in which agency, and everything which follows from agency, is attributable to human beings is where the development has arrived but not where it began.

The idea is that we can see what consciousness is built from – it is built out of subjectivity, a special sort of evolutionary product – but can see that it is built in a gradualist way and closely tied to basic features of life.

Subjects are explicable products of animal evolution. In bringing the biological story to bear on the philosophical problems, we see interesting possibilities of dissociation between subjectivity-relevant properties, and may have to grapple with a graded conception of subjective experience and its origins.

Peter Godfrey-Smith: Materialism, Subjectivity, and Evolution

In this essay, Godfrey-Smith is primarily concerned with the possibility of a naturalist and materialist explanation of subjects in support of which he sketches a picture of the biological evolution of subjects.

A picture we might start out from recognizes two concepts: subjectivity and agency. These have different emphases – subjectivity is more a matter of the input side; agency involves the output side. But from a biological point of view and perhaps others, these are largely correlative and complementary capacities; sensing and action coevolved, and each gives the other its point. Initially, I’ll think of subjects as having a pair of features: (i) a point of view on the world, and (ii) an agenda. Subjects act in ways that reflect both.

That is, what distinguishes subjects is subjectivity; subjective experience is the experience of a subject. In Thomas Nagel’s formulation, there is a something it’s like to be about subjective experience. This idea can be fleshed out both from the input side as a perspective on sensory perception and point-of-view and from the output side as the feel of and evaluation of experience.

Even the simplest prokaryotic organisms have the capacity to detect changes in their environment and coordinate their movement in response. However, it is the distinctive multi-cellular systems in animals that are most relevant here; more specifically, sensory structures such as eyes that make it possible to track the environment at a macro-scale, effector systems of muscles that allow co-ordinated movement in response, and a control system, a nervous system, that links the two together.

These systems appeared very early in the evolution of life on earth. Nervous systems are present in corals, jellyfish and anemones as well as in animals with bilateral body plans, an evolutionary divergence that occurred maybe 600 million years ago in the Ediacaran period.

In the Cambrian period, beginning around 540 million years ago, sophisticated sensory and motor systems evolved and animals started to interact with each other, often in the form of predation. Animals in three groups have developed complex active bodies capable of rapid motion, sensing at a distance and the capacity to manipulate objects. These are cephalopods, marine molluscs that include octopus, squid and cuttlefish; arthropods, both marine arthropods such as crabs and land arthropods such as insects and spiders; and all the vertebrates, fish, amphibians, reptiles, birds and mammals.

Connectedly, these forms are also characterised by image forming eyes, sensory integration, compensation for reafference and instrumental learning.

Image forming eyes support the genuine viewing of macro-scale objects. The integration of diverse sensory channels generates the subjective point-of-view. Compensation for reafference is the name given to an animals ability to track the effects of its own actions on what it senses and therefore indicates the beginnings of a distinction between self and other. Finally, instrumental learning is the capacity of an animal to track the link between actions and their positive or negative consequences and to adapt its behaviour accordingly. It is through instrumental learning that an animal’s agenda is bought within internal control.

Intuitively, you might think that increasing sophistication on the input and the output side would go together, but Godfrey-Smith suggests that this isn’t always the case. Insects may be stronger on the sensory side, especially those that can fly, but they typically live very short, very routine lives. This makes sense if the adult stage of an insect’s life is lived as little more than a disposable reproduction machine.

On the other hand, gastropods, molluscs such as slugs and snails, have much more limited sensory capabilities but there is some evidence to suggest instrumental learning in, for instance, the way that a damaging event appears to lead to a state of negative readiness, a kind of corporeal wariness. Such a capacity would make sense in an animal such as a sea-snail that is vulnerable to damage but also relatively long-lived and which has the ability to repair itself.

This kind of differentiation is relevant to the question of whether the evolution of sentience is gradual or organised around thresholds. Does subjective experience gradually feed in or does it require a threshold to be reached?

The threshold model is that there is a lower bound in complexity which must be reached in order to generate sentience. The question becomes – what features suffice for an animal to be sentient. The gradualist model, on the other hand, tries to account for how some organisms have internal workings that are experiential.

In this domain, much is tentative, but the gradualist model is supported by the distribution of features and the existence of every possible intermediate and partial case, both on the sensory side and the evaluation side. It seems, for example, there may be some level of reafference compensation, and therefore the glimmer of a sense of self, even in nematode worms.

How does this link to philosophical arguments around the concepts of naturalism and materialism?

Godfrey-Smith is describing biological features that start to close the gap between the physical and the mental. These features give an entity a point-of-view, a sense of self versus other and endow events with a positive or negative value. They are not just kinds of complexity. Once we have subjects, we have subjectivity.

The idea is that with a filled-out story of this kind – one that I have only sketched here, but which I think we can glimpse – there is no extra question of whether there’s something it’s like to be an animal of these sort. Once we have subjects, subjectivity comes along. The way it feels to be a system of this kind comes along. These biologically established points of view are occupied, not merely parts of the world’s layout.

The development of specifically human capabilities such as the use of language and the ability to understand conceptual meaning requires further evolution but what this suggests is that that subsequent evolution has less to achieve than might be supposed. It doesn’t have to establish macro-scale imaging of the environment, it doesn’t have to generate a point-of-view or separate the self from the environment, and it doesn’t have to learn how to learn. These features and capabilities are all inherited, at least in basic form, from the evolution of sentient beings.

The narrower the gap, the easier it is to bridge.

One way of interpreting this is that if we can show the explanatory capabilities of natural science moving up the evolutionary scale towards human experience, the possibility of a naturalist explanation for specifically human capabilities becomes more likely and a materialist ontology becomes more plausible.

I think this is how Godfrey-Smith reads it. As he remarks, if materialism is true, specifically human capabilities must in some sense be an illusion.

There is an appearance of separability, and you either explain it as a probable illusion, or not. Everything else is secondary, or less than secondary.

Underpinning this idea is the intuition that a materialist ontology and a naturalist epistemology and methodology go together and are mutually supporting. Metaphysical gaps and epistemological gaps can be correlated and therefore metaphysical gradualism implies epistemological continuity.

In my view both materialism and naturalism are problematic concepts. I will consider them in another article but I think that, rather than being mutually supporting, naturalism and materialism trip over and undermine each other. This isn’t typically considered a problem because for the most part materialism is just the background metaphysical assumption to the project of naturalism and usually isn’t itself considered to be “in play” as it were.

Having said that, I don’t believe we have to adopt any form of dualism as an alternative. Materialism is a form of monism, but not the only possibility. Just as we can see that every biological entity is also a physical system, without thinking that physics is sufficient to explain living organisms, we can also see that every introspectively intelligent being is also both a biological entity and a physical system without supposing that biology and physics are adequate to the explanation of human experience.

If this is right, then we can still appreciate how much the specifically human experience inherits from animal sentience, and the narrowness of the differences between human beings and some animals, without becoming committed to either a materialist ontology or naturalistic epistemology.

Peter Godfrey-Smith: Materialism, Subjectivity, and Evolution 2017,