Basic HTML Version

consequence of the ways in which the system responds robustly, even in the face of deletions
of internal components, to a change in the environment. This kind of robustness is one feature
of emergent properties. Emergent properties do not depend merely on the aggregation of
components, but are instead dependent on the system’s dynamic responses to the absence of
components as well as the external and internal environments.
Honeybees and emergence
In another example of emergent properties, social insects, there is evidence of downward
causation. In a honeybee colony, what causes a particular bee to fly out of the nest to forage
for nectar at a particular time? The answers that would be expected from a reductive account
would focus on the individual, for instance that the bee has genes that code for a certain
threshold of response to stimulus. One bee might be a nectar forager and another a pollen
forager, and their different genetic makeups explain why one forages for nectar and the other
does not. That would be an example of a bottom-up explanation.
This is a piece of the story, but it is not going to provide the whole story about why a
particular bee goes out and forages for nectar at a particular time. It turns out that the
behavior of individual bees is responsive to a property of the colony as a whole. The property
is the amount of nectar already stored in the colony. The colony, the higher-level system, has
a property that is not that of an individual bee, namely how much nectar it has stored, which
influences the behavior of the individual bees to go out and forage for nectar and bring it
back to the colony.
How does the colony obtain nectar? The nectar is produced by all the individual bees
going out and foraging for nectar and bringing it back to the colony so that it can be stored.
There is a sense in which this higher-level property is generated by the behavior of
individuals, but at the same time it influences the behavior of individual bees. How does it do
this? There is no centralized account of how much nectar is stored and nobody directing the
bees to do what they are doing. There is, however, a system of self-organization whereby the
information about what is stored in the colony is available to the bees in terms of how long
they must wait in order to unload the nectar they bring in.
The explanation of how bees know how much nectar is stored in the colony depends
on the higher-order property, which is itself generated by the behavior of individuals. The
result is a complicated feedback system in which the emergent property of the amount of
nectar that is stored plays a crucial role. An individual bee flies out to collect nectar and
returns to the colony. Other bees waiting at the colony unload the nectar and then fly around
internally to find an empty cell in which to deposit it. While they are doing that, another bee
lands and has to wait for the unloading bee to return. The time it takes for the unloading bee
to return is directly correlated with how many empty cells there are: if there are many empty
cells, the wait is short; if the colony is full of nectar, it takes a long time to find an empty cell.
Thus, the length of time an incoming bee must wait reflects a higher-level property and this
influences the probability that the bee will return to foraging. This is an example of
downward causation: the higher-level property is changing the behavior of the components,