Emergence: An Introduction

What is emergence? Answering this question is no small task, and making progress on doing so is one of the key concerns of the Durham Emergence Project. Roughly speaking, to claim that something is emergent involves asserting something about the relationship between that thing and it’s more fundamental parts – for instance, that whilst that thing is dependent on its parts (that is, could not exist without them) it is also novel with respect to them; it is something new and distinct. As we will see, exactly how we should elaborate on such claims is a matter of much debate.

Emergence is a controversial topic for philosophers. Philosophers tend to divide the world into fundamental objects (‘entities’) and properties. The category “fundamental” is often guided by what scientists view as fundamental, so the laws of physics are typically taken to be fundamental whereas the laws of society are not. We may say electrons are fundamental particles because they are indivisible and not made up of any other entities, whereas tables are composed out of components, for instance atoms.  When we bring together some of these fundamental elements some argue that we have brand new properties or entities. The sum of parts is both more than and different from the parts considered alone. To some this is a contradiction in terms, since either we have something new that cannot be described in terms of the fundamental entities, hence it itself must be fundamental, or we have something that can be described or grounded in terms of fundamental entities and properties and hence is no more than a result (‘resultant from’) of the natures of those entities and the relations in which they stand.

One way of conceptualising emergence is as a failure of reduction. The core notion behind a ‘reduction’ is that some relatively complex phenomenon can be explained in terms of some more simple phenomenon. Common examples often given are that, for instance, the seemingly continuous and ‘fluid’ motion of liquids is in fact just the ‘tumbling about’ of loose molecules—and as such, is not really all that different from other familiar macroscopic phenomena, such as the tumbling of glass marbles as they’re emptied from a jar.

Descartes argued that animals could be reductively explained as automata such as the digesting duck.

As new sciences are developed and they probe deeper into reality exploring more the world further and further from our everyday experience, we typically see more reductions proposed. Often these new theories or entities seem to then specify what higher level entities are composed of. So for instance it is argued by many that a fundamental theory like quantum mechanics grounds everything that the science of chemistry has to tell us, since in some sense the units of chemical investigation, molecules, are made of more fundamental particles governed by the laws of quantum mechanics. In contrast to this, emergentism holds that there is some sense in which appeal just to these lower level phenomena will be inadequate—at least some macroscopic phenomena cannot be explained away by reduction. The challenge for supporters of emergence is to say what is added as we move up levels of reality, and in what way higher level phenomena are not merely resultant from the interaction of lower level entities.

The Durham Emergence Project brings together physicists, philosophers of science, philosophers of mind and metaphysicians. Each group will approach the topic of emergence from within their own discipline but the groups will work in collaboration to bring new insights into each other’s fields. We will investigate new potential cases of emergence from physics, whilst considering how emergence might be categorised in the broadest philosophical terms. Our aim is to produce a more clear understanding of how different kinds of emergence may be philosophically defined and defended and what evidence (if any) there is of their existence.

Categories of Emergence

Epistemic or Ontological Emergence?

Even if it’s granted that there are emergent phenomena, not all types of emergence will be the same. One way of dividing up different types of emergence is between that which involves phenomena that emerge because of the way humans represent or gain knowledge about the world (epistemic emergence) versus the emergence of genuinely existing new entities, powers or properties (ontological emergence). Epistemic emergence is relatively uncontroversial, as most scientists and philosophers believe that it is sometimes convenient to describe (or ‘represent’) the world in certain ways, and depending upon the particular description (or ‘representation’) we will have different entities. For example, a table might be considered to be a solid object in one representation, and to be a collection of atoms arranged in empty space in a particular way in another. Which representation you use will depend on which is most useful for the task at hand.

Ontological emergence is a different notion. It proposes that the higher level will contain entities which are just as real as the lower level entities. This ‘realness’ might manifest itself by the higher level being able to act downwards to affect the lower level or in a failure of explanatory reduction. Often scientists explain events in the world by invoking unseen entities, such as explaining how your TV works by discussing electrons. One way of understanding what these explanations which involve unseen entities like electrons mean is to argue that unless these entities are really there we could not explain such events (this is an example of a ‘realist’ stance). One way of arguing for ontological emergence then is to employ the same logic: if we can’t in principle explain a phenomena without having to use entities or properties that aren’t fundamental and we take the unseen fundamental level seriously, we have good reason to take the non-fundamental properties just as seriously.


To be a realist about some particular domain or class of objects or entities is to hold that those things exist in some mind-independent way: even if there were no humans in the universe, those objects or entities would exist in just the way they do. That is to say, those objects’ existence does not depend on the beliefs, conceptual schemes, linguistic practices, or preferences of humans. Realism is often contrasted with a pragmatist approach, which holds that we should simply treat some things as if they really exist if it is useful to do so, but should refrain from taking a fully realist attitude towards them. Questions regarding realism arise in a huge variety of philosophical contexts, not just in metaphysics and the philosophy of science. For instance, the moral realism debate questions whether or not notions such a ‘good’, ‘bad’, ‘right’ and ‘wrong’ ought to be understood in a realist manner; and in aesthetics it is debated whether properties such as ‘beauty’ are real, or just “in the eye of the beholder”.

Strong or Weak Emergence?

Another way of dividing up classes of emergence is into strong and weak emergence. Sometimes this division follows the classes of epistemic and ontological emergence, so that strong emergence implies emergent entities exist and weak emergence implies that apparently emergent entities are merely resultant from the way humans gain knowledge of the world and represent it. This identification is not always so easy however. For instance Wilson (2012) argues for weak ontological emergence based on the elimination of the number of degrees of freedom a system has. The idea here is that we do not have new powers at the higher level but we also have an object, described by the number of degrees of freedom, that is not identifiable with any of its parts or combinations of its parts. Equally, Huneman (2012) has argued that there are objective ways of defining epistemic emergence in terms of computability. As such even though we have no new entities we would not have emergence as an artefact of human perception. Strong and weak emergence are best thought of then as categories that can sub-divide epistemic and ontological emergence even further. In strong ontological emergence we have a commitment not to new entities as such, but to new agents at a higher level which can exercise powers. In weak ontological emergence we have no commitment to new powers or agents but we have a class of objects that cannot be identified with any way of summing up their constituent parts. In strong epistemic emergence we have no new agents or entities but we do have a robust objective reason why the world needs to be partitioned in certain ways. Whereas in weak epistemic emergence all we are doing is representing the world in some way or other for calculational ease

Emergence as Aggregation?

One familiar approach to emergence involves aggregation. A complex system may be emergent if in some sense it transcends its parts, in the sense that it has properties that cannot fully be explained by reference to their properties, considered individually or in combination. But it is not clear how this aggregative conception covers all the candidate cases of emergence. Even if it works well for some cases in physics, it is unclear how it should apply in the philosophy of mind. What is it that is aggregated to generate the bearer of a mental property? Mental properties are not borne (for instance) by mere aggregates of neurons. Rather, mental properties are maintained by dynamical interactions within systems of neurons that have particular kinds of causal history that support a mental life. In fact the metaphor does not seem to do justice even to physics, which also seems to require particular causal histories in some of its candidate cases of emergence (preparation of spin-paired photons in singlet states, for instance).


Huneman, P. (2012). Determinism, predictability and open-ended evolution: lessons from computational emergence. Synthese, 185(2), 195-214.

Wilson, J. (2010). Non-reductive physicalism and degrees of freedom. The British Journal for the Philosophy of Science, 61(2), 279-311.