[This is an invited guest post by David A. Haig (Harvard, Biology), which may, but need not, be read as a comment on this post--ES]
Chickens come from eggs
But they have legs.
The plot thickens:
Eggs come from chickens,
But have no legs under 'em.
What a conundrum!
Eggs produce chickens and chickens produce eggs. This is a fact universally acknowledged, but when a respondent at a scientific meeting raises the question “Isn’t this an example of the chicken and the egg?” what is usually being offered is an inversion, “you say X causes Y but you have it the wrong way round, Y causes X.” It is considered obvious that X cannot both be a cause and an effect of Y.
This ‘natural way of thinking’ expresses a common confusion that arises from the extrapolation of the logical properties of token-causation to type-causation. If X and Y are token events and X occurs before Y, then X can be a cause of Y but Y cannot be a cause of X. It is an empirical question whether a particular chicken developed from a particular egg. But there is no logical contradiction in claiming that events of type X could be both causes and effects of events of type Y because tokens of X can occur both before and after tokens of Y. The claim ‘events of type X are causes of events of type Y’ and its inversion ‘events of type Y are causes of events of type X’ could both be true, both be false, or one be true and the other false. In this regard, it should be noted that when scientists formulate scientific ‘laws’, these explicate regularities between types of events. The ‘laws’ are concerned with token causation only in so far as token events are exemplars of general patterns.
Life is a recursive process in which progenitors were progeny of progenitors who were themselves progeny, based on the recursive process by which copies of the DNA double helix are themselves copies of copies. By processes of natural selection, this reproductive recursion is the source of the purposiveness of living things. A genetic variant that causes a variant behavior will be copied more frequently if the behavior indirectly enhances copying of the genetic variant. Behaviors, and their consequences in the world, are responsible for which genes are copied. The causes of why this gene rather than its variants is present in a species is to be sought in the gene’s effects in the world. Genetic variants arise by ‘random’ mutational processes but these variants are then subject to the complex environmental sieve we call natural selection. And that which is sieved is subject to further mutation and resieving by the environment. By this process, information from the environment is incorporated into the genetic material.
Most of my biological colleagues believe it is intuitively obvious that genes cause behaviors rather than the other way around. But token behaviors occur both before and after token expression of the relevant genes, and a complete causal account of why the genes that we see have the sequences they possess, rather than other sequences that have existed but have failed to leave descendants, will include the effects of these genes in the world. A gene’s effects (where a gene is considered as a historical lineage of gene tokens) have a causal role in the gene’s survival of the sieve of environment. It is no more than a prejudice to consider the claim that genes determine the properties of behaviors to be more ‘scientific’ than the claim that behaviors determine the properties of successful genes.
Natural selection is a purposeless process that produces beings with purposes. Natural selection is not teleological but it produces living beings for which teleological language is entirely appropriate. For innate behaviors, a token behavior is performed because many past tokens of the same behavior were associated with effective action in the world. The behavior exists for this purpose. Purposeful action is the ‘anticipation’ of future effects. For innate behaviors, what has worked in the past is ‘anticipated’ to work in the future. And what has worked in the past is the use of information gained from the environment in a given life to guide behavior in that life.
Organisms have evolved to learn and perform behaviors for reasons that are specific to each individual life. The ‘anticipation’ of conditioned reflexes is very much like that of natural selection: what has worked in the past is ‘anticipated’ to work in the future, but we also have abilities to simulate outcomes of different choices of behavior and to choose the behavior with the most desirable predicted effects.
Many choose to restrict the use of intentional idioms to choices based on simulated outcomes but this is a choice of definition. I prefer a more expansive definition of purposes, intentions, and final causes, both because this is the way we actually use language (witness a molecular biologist talking about codes, transcription, translation, and RNA editing) and because the expansive definitions emphasize continuities in nature rather than create artificial boundaries (and a haven for dualism).
A fascinating post. Lots to wonder about here. Not least, "information from the environment".
But it's worth noting that you felt the need to put "anticipated" in scare quotes. Doesn't this suggest that you think that intentionally-loaded terms in this context should be viewed with skepticism - that there's a worry about how such language is used around biology?
There has been, as you know, some controversy about whether evolutionary biology has the machinery to explain representational content. Won't the preference for an expansive reading of purposes either beg the question, or lead to unhelpful equivocation?
Posted by: ajkreider | 07/27/2018 at 04:59 AM