Teleology is the technical term for goal-directedness, especially when describing living systems. Teleology has been problematic in the sciences because of the amount of hand-waving that teleology has historically allowed. From the outside, it is difficult to tell if something happened because it was intended or if it just happened to be beneficial. Determining the precise goal can be problematic, even if an action is goal-directed. It is easy to construct a story about why an organism does an action, but how do we ascertain whether this story is true? When are attributions of teleology science, and when does it degenerate to mere invention? Additionally, the lack of ability to measure goal-directedness has often placed teleology in the realm of storytelling instead of science.

However, it has long been recognized in biology that organisms are so abundantly goal-directed that it is impossible to reason about their activities effectively without talking about teleology. Nearly every action that an organism does is for something. However, in the 19th and early 20th centuries, biologists aimed to reduce teleological actions to non-teleological causes (a viewpoint often termed reductionism). Organisms were viewed as being only collections of atoms, and atoms were considered mere passive conduits of the equations of physics.

This viewpoint was always known to be problematic for understanding behavior. Still, it was thought this was a limit to the human ability to measure and calculate the low-level phenomena precisely. Thus, biologists were comfortable analyzing organisms from a high level teleologically but largely believed that they were, at the bottom, mere collections of blind particles interacting. Therefore, the field of biology viewed itself as more successful when it did not invoke purposes than when it did. 

This view especially permeated the study of evolution. The modern theory of evolution came into its own in the heyday of this reductionist thinking, and thus, teleology was excluded from evolutionary theory essentially by fiat. While there were some experiments that people often pointed to as justification, ultimately, these didn’t support the inferences people made from them. Additionally, while natural selection was thought to be capable of causing large-scale directed evolution, many of the capabilities attributed to selection relied on the very teleological language that biologists were seeking to eliminate. 

As a mechanism, selection itself was thought to be sufficiently efficacious to handle the job of evolution. Fisher’s Fundamental Theorem of Natural Selection was supposed to show that selection was perpetually increasing a population’s fitness mathematically. However, this confidence was misplaced. Fisher’s Fundamental Theorem actually said very little about evolution itself, as the theorem was based on non-mutating populations. This did not bother Fisher, as he assumed that mutations were well-distributed between beneficial and deleterious mutations. However, more recent studies, based on empirical studies of mutations, have shown that Fisher’s optimism was not well-founded. Additionally, even as far as it goes, the theorem says very little about selection improving real organismal fitness as it is usually conceived. Instead, it relies heavily on an equivocation about what fitness means for organisms.

No Goals in Unguided Evolution

In the move to understand evolution without teleology, biologists came up with a more nuanced vocabulary of teleology. Causes based on purpose were still considered teleological, but another term was developed as well — teleonomy. A teleonomic process is a goal-directed process whose purpose is encoded in some form of mechanical system or code. The idea is that organisms exhibit teleological activity because their biology (especially DNA) mechanizes this purpose.  However, biologists were adamant that the process of evolution itself was not teleonomic — there was no system controlling the direction of evolution towards any goal. Biologists such as Ernst Mayr were trying to cut the chain of teleological causation so that organisms could exhibit teleological activity because of their biology, but the biology itself has a non-teleological explanation. There is nothing about the term teleonomy that would logically prevent it from being applied to evolution, but it was the emphatic belief of those who developed the modern synthesis in evolutionary theory that evolution was not in any way teleological or teleonomic.

At the turn of the 21st century, many biologists have recognized the need for a less reductionistic biology that takes purposes and goals (and the organism as a whole) more seriously as centers of causation.  In “A New Biology for a New Century,” biologist Carl Woese  says, “The machine metaphor certainly provides insights, but these come at the price of overlooking much of what biology is.” If we return to looking at biology as more fundamentally teleological, then we have to revisit previous questions we were trying to avoid.

So how can scientists (especially biologists) utilize teleology non-reductionistically while avoiding the non-rigorous handwaving that has pervaded teleological explanations in the past?

The first task is to address what properly constitutes a goal. Is it enough for the proposed goal to be beneficial to the organism? Is it sufficient for the organism to be more likely to perform a particular action given a starting point?  These criteria for goals are helpful in some ways but, ultimately, are too subjective and are not sufficiently specific to identify goals objectively. Thankfully, philosophers of science have been working on this problem in recent decades. One of the more promising criteria for objectively determining the goals of an organism was proposed by Mossio and Bich. For them, an objectively identifiable goal is something that is involved with maintaining the organism as a distinct, coherent causal unit while simultaneously requiring energy/effort to achieve. While an organism may (and likely does) have other goals, this is at least one way of objectively identifying goals for organisms.

So, if we now have a way of identifying goals, how do we measure the amount of directedness an organism has towards a goal? A development in computer search optimization theory can help with this. In computer search, “active information” measures the amount of information an algorithm contributes to a search. Essentially, if a search algorithm is not specifically aligned with the search space, then the expected value of the search algorithm is equivalent to a random search. Therefore, if you compare the outcome of a given search algorithm to the result of a random search, you can measure how much information the algorithm adds (or detracts) to the ability to find solutions.

While this was originally applied to search algorithms, the usefulness of this measurement technique is fairly broad. For biological organisms, one can use active information to measure intentionality by comparing outcomes of organisms achieving objectively identifiable goals to what the result of similar non-teleological objects and/or random walks would achieve in similar circumstances. While other measurement techniques have been proposed, active information has been shown to have a lot of beneficial mathematical properties stemming from its origin in information theory.

So, armed with an objective way of identifying goals and a method of measuring goal-directedness, can any of this be utilized within evolutionary theory? It turns out that evolution is much more teleological than has been historically supposed. Not only has the prior evidence for the non-teleology of evolution mainly been overturned, but new research has increasingly focused on the teleological and teleonomic causes that underlie much of what shapes the direction of evolution. The successor to the modern synthesis (i.e., neo-Darwinism), known as the “extended evolutionary synthesis,” can largely be described as incorporating teleonomic causes into evolution. A recent volume published by MIT Press is entirely about new discoveries that are being made about the role that teleonomic causes play in evolution.

What Would Teleonomy in Evolution Look Like?

One interesting example is how E. coli responds to starvation. Typically, E. coli can’t process arbutin as a food source.  However, if E. coli is grown in the presence of arbutin and starved, a process causes a specific DNA sequence to be inserted upstream of a previously inactive gene (glpFK) that turns the gene on. In other words, specific conditions trigger a specific mutation that allows the organism to process the food source in the right conditions. The mechanism by which this happens is still under investigation, but current  models propose that starvation causes the transcription factor that is usually blocking the insertion site to clear away while another transcription factor made in the presence of arbutin causes the DNA to twist in a way to expose the insertion site to activity.

Another interesting example is the bacteria Neisseria. It has a gene that codes for pilin, an outer coat protein. Depending on the environment, different pilin genes would be beneficial. Rather than having to randomly mutate through different DNA sequences and hoping to find one that alleviates current selection pressures, Neisseria has a set of alternate DNA sequences for its pilin gene encoded in a series of pseudogenes. Therefore, it varies its outer coat through pre-encoded variants rather than random sequence changes.

So, where does this teleonomy originate from? Can this teleonomic behavior of evolution simply be a byproduct of non-teleological forms of evolution? Information theory suggests that this is not likely. Ultimately, searches only work when they are biased in such a way that makes finding solutions possible. These mechanisms may have evolved from previous mechanisms, but that would require the organism to start with more information, not less. This means that some form of teleology more than just teleonomy is needed to get evolution as we see it today to work, whether that comes from inside the organism or outside of it.