To quote from the Stanford Encyclopedia of Philosophy:

Here Aristotle recognizes four types of things that can be given in answer to a why-question:

  • The material cause: “that out of which”, e.g., the bronze of a statue.
  • The formal cause: “the form”, “the account of what-it-is-to-be”, e.g., the shape of a statue.
  • The efficient cause: “the primary source of the change or rest”, e.g., the artisan, the art of bronze-casting the statue, the man who gives advice, the father of the child.
  • The final cause: “the end, that for the sake of which a thing is done”, e.g., health is the end of walking, losing weight, purging, drugs, and surgical tools.

My assumption has always been that modern science concerns itself mostly with first three causes and that the final cause is considered out of bounds. Apparently Francis Bacon put both formal and final causes into the metaphysical realm:

For as we divided natural philosophy in general into the inquiry of causes and productions of effects, so that part which concerneth the inquiry of causes we do subdivide according to the received and sound division of causes. The one part, which is physic, inquireth and handleth the material and efficient causes; and the other, which is metaphysic, handleth the formal and final causes.

But I was surprised to read that biologists especially (including Darwin himself) have begun to tackle even Teleology, the Fourth Cause. Is this appropriate in Modern Science?

  • 1
    Of course, I ask now because I want a chance at a free book, but I've long wondered why teleology began to fall out of favor ironically around the same time Aristotle began to be rediscovered in the West. Commented Dec 14, 2011 at 21:27
  • 2
    I would disagree with the idea that biology has anything to do with causes. Evolution is not a directed process. I can program a "genetic algorithm" that will do a massive number of different random things, pick the best, and then modify that in a massive number of different ways, pick the best, and repeat. There is a progression towards the optimal solution. The random number generator has no grand plan and no goal. It doesn't want to make a perfect solution, it doesn't want to make a terrible solution. It simply spits out random numbers. What do you mean by the statement? Commented Dec 15, 2011 at 2:29
  • @Keller: If you read the Wikipedia entry I linked to or this one, you'll see that evolutionary biologist make statements that read as teleological. Any time an author uses the phrase "in order to" they are making a teleological statement whether they mean to or not. Presumably they don't mean to for reasons such as the one you suggest. (It's interesting that you used the phrase "progression towards the optimal solution", which strikes me as teleological too.) Commented Dec 15, 2011 at 17:02
  • 1
    I did read it. I agree that evolutionary biologists make those statements. I don't believe that they are addressing ultimate questions of intent. My action has a goal. The fourth cause of the computer spitting out random numbers is that I wanted it to produce something. The program is not the fourth cause of the data. I produce the program with the intent of having it spit out random numbers until it gets a set that is good enough. The data doesn't have a different fourth cause generated by the program. A dog doesn't have a cause generated by its DNA, though it approaches optimal DNA. Commented Dec 16, 2011 at 1:56
  • @Keller: Sorry. I must have misunderstood your earlier comment. At any rate, while genetic drift doesn't have a purpose, the process of natural selection is sometimes said to have a purpose of selecting mutations that best provide characteristics that facilitate reproduction. The anthropic principle verges in that direction sometimes too. I've asked a related question which might help you understand where this question comes from. Commented Dec 16, 2011 at 17:03

7 Answers 7


This is a huge question actually and one of considerable current research. I would recommend you start with the Stanford Encyclopedia article: http://plato.stanford.edu/entries/teleology-biology/

Another excellent place to start researching the issue if you'd like to delve deeper is the work of a very philosophically sophisticated biologist named Ernst Mayr. I'd particularly recommend his essay The Idea of Teleology in the Journal of the History of Ideas, 53(1), 1992.

Another place to look if you want to delve deeper is the work of William Bechtel, who does contemporary philosophy of biology/neuroscience. His work is on the notion of function in biology. I think function is the key--if you want a scientific description of functional systems like living organisms and their parts, then you have to give such a description in functional terms, which means you have to have teleology. e.g. "the kidneys are for filtering the blood". You can see a slideshow of his with some interesting info about teleology and functional explanation in contemporary biology here

My view, which according to the author of the SEP article, is widely shared among contemporary philosophers of biology is that: "Many contemporary biologists and philosophers of biology believe that teleological notions are a distinctive and ineliminable feature of biological explanations but that it is possible to provide a naturalistic account of their role." And I think Bechtel is right to say that functional analysis is the way to offer a naturalistically acceptable way of getting function talk going.

Speaking historically, I think this functional systems approach to teleology is actually probably a lot closer to Aristotle's original conception that some of his contemporary critics recognize. The debates about evolution and design of the 19th century were conducted in terms that assimilated the Aristotelian idea of teleology to divine providence and God's guiding events towards his purposes, but I don't see any of that kind of view at work in Aristotle himself.

  • Welcome to Philosophy--Stack Exchange! You've given me a lot of information to chew on. It sounds as if teleology is inevitable in biological explanations because so many processes are required in order to sustain life. Life or, more accurately, the propensity for DNA to propagate, is the final cause that biologists have in mind. Since genetic information self-duplicates, that duplication is the purpose of all biological functions. Is that a fair summary? Commented Jan 7, 2014 at 19:38
  • It's more than just reproduction. It's also how does the cell maintain homeostasis? How does it repair itself? You answer those questions by citing mechanisms in the cell that perform those functions.
    – user5172
    Commented Jan 7, 2014 at 19:50
  • The 'mechanisms' in cells are driven by an innate form of primal intelligence. The messaging function which instucts the zygote to form a complete human being can only be compared to and perhaps far outstrips the intelligence which allows us to hold this discussion. That is the logical endgame to where biology has been headed since Oparin in the last century.
    – user37981
    Commented Oct 10, 2020 at 3:52

Modern science doesn't consider Aristotle's final cause to be a cause. The modern meaning of the word cause is simply different from the meaning of the word as used by Aristotle. However, modern science still considers describing "relevant ends" as providing valuable insight. This insight will typically be unrelated to the origin of the thing, but modern science is well aware that there may be more important questions than just the origin.

  • This is also not true--- the "final cause" is fine when you are talking about intentional systems. Why are male peacock's tails longer today than 10,000,000 years ago? Because the females wanted future tails to be longer. This is a form of final cause, the female choice is intending a future outcome. Final causes are everywhere in evolution, and are not anathema to science if they are implemented by material causes.
    – Ron Maimon
    Commented Apr 22, 2012 at 18:33
  • @RonMaimon I answered this question back then, because I had already investigated that question before. If you look at the four causes, only the "efficient cause" resembles the way we use the word "cause" today. But the "final cause" seems to imply some sort of teleology, so people worry more about it than about the other "slightly off" causes. My answer just tried to clarify that even modern science sometimes investigates "relevant ends" (what Aristotle called "final cause") without subscribing to any sort of teleology. Commented Apr 23, 2012 at 8:47
  • I see, you're talking about the use of the word "cause". That's true, you're right--- usually we don't call teleological cause "cause", but "epiphenomenon".
    – Ron Maimon
    Commented Apr 23, 2012 at 16:30
  • 2
    Why does the female peacock respond to feathers of a certain kind? Certainly not because she "wanted" future tails to be longer. That's absurd. There isn't necessarily a particular "reason" she should respond to a particular peacock other than by virtue of her biological constitution (or why a male should respond only to females with all their feathers). That the effect may be survival of the species is also not a matter of intent (peacock feathers can attract predators just as well). As Gilson notes, while modern science doesn't concern itself with finality, its methods do not contradict it.
    – danielm
    Commented Nov 21, 2012 at 11:57

Teleology is, generally speaking, within the domain of theology; to speak of a final cause implies a subject who is intentionally causing the action in question, which in the global sense would apply to a god of some sort. (Naturally, teleology with regard to human actions is in the domain of psychology, but that's not what interests us here.)

The history of modern science, on the other hand, has been predicated (for the most part) on a growing assumption of a naturalistic hypothesis, where recourse to a deity is off-limits; the principle of parsimony implies Occam's Razor, so positing an additional entity to serve as the subject of a telos is frowned upon, as it is more parsimonious to just assume an unmotivated process. So, in general, the more modern a scientist is, the less likely they are to partake in teleological deliberation.

  • This is false--- you can assume that natural phenomenon in biological evolution have a teleology without assuming that there is intervention of external supernatural entity. It might just be a law of nature that evolving computations become more complex and self-aware--- it is certainly true of life on Earth, and human social development up to this point.
    – Ron Maimon
    Commented Apr 22, 2012 at 18:31
  • You can assume that, but you have no particular reason to-- life on earth has gone through many waves of expansion and contraction; I'm sure the dinosaurs were certain that it might be a law of nature that creatures would keep getting bigger. Commented Apr 23, 2012 at 6:59
  • I don't agree with either of you, but I agree that finality is not theological per se. After all, Aristotle's metaphysics certainly concerns itself with final causality, even going so far as to say that the organization of animals is impossible to explain without finality. But of course, he does not adhere to a mechanistic metaphysics, in which case teleology must be externalized just as the finality of an artifact is external to the artifact itself. In that case, even the often vague idea of physical law can't be teleological unless it rests in the mind of God.
    – danielm
    Commented Nov 21, 2012 at 12:13

Haldane supposedly famously quipped "teleology is like a mistress to a biologist; he dare not be seen with her in public but he cannot live without her".

Teleology finds expression in Lamarckism - and such accounts of evolution are relegated to being added to Kiplings "Just so" stories. Darwin solved the problem for mechanical science by positing that 'natural selection' was the causal mechanism which life evolves (no teleology involved). Darwin theory finds a parallel in Pavlov; there is nothing the dogs can do to get their meat powder (indeed Pavlov had to prod the dogs to keep them awake). This is unlike Skinner's rats, which were only rewarded for the consequences of their behaviour.

Non-consequentialism (non-teleological) evolution was partially amended in the 20th century by the notion of co-evolution. Bateson, for example, suggested that the grassy plains evolved in response to the evolving teeth and hooves of the horse and other ungulates. Thus, although the animals were not directly selecting their own evolution, by changing the turf they were indirectly changing themselves by creating an environment that some of their number thrived in. It is the context which evolves, and the moral is simple, we create the environment that will select our children.

Teleology was further brought out of the closet, or rescued from metaphysics where Bacon had relegated it, with the definition of life as 'autopoietic' - the notion that all living things are organised in such a way that they know how to produce themselves. From the simplest forms they know how to go out and harvest the materials to produce themselves.

Now that organisation is mechanical (billiard balls hitting each other) but it results in an organism that is goal directed (purposeful or teleological) - the bacteria detects sucrose richness ahead and swims to it. Thus we find teleology no longer shyly hiding in the biologists closet, but arm and arm with him and her head held high.

  • Pavlov and Skinner's dreary views on human nature and their stultifying application of behavioral animal replication into human psychology is the single most powerful impediment to rectifying the unacceptable comparisons of humans to the much lesser creatures.
    – user37981
    Commented Oct 10, 2020 at 4:07
  • Welcome to SE Philosophy! Thanks for your contribution. Please take a quick moment to take the tour or find help. You can perform searches here or seek additional clarification at the meta site.
    – J D
    Commented Oct 12, 2020 at 16:40
  • @CharlesMSaunders Now now, behaviorism has its flaws, but to blame doctrine and not critical thinking for poor conclusions untenably exculpates thinkers. :D
    – J D
    Commented Oct 12, 2020 at 16:42
  • I agree NickDrury that teleology is highly useful and shouldn't be scorned! +1 For interested readers: SEP: Stanford Encyclopedia of Philosophy.
    – J D
    Commented Oct 12, 2020 at 16:43
  • @J D- Marvelous move! Did you learn that tap dance at the Arthur Murray studio? But seriously, anyone who believes that humans are simply stimulus/response automatons needs immediate psychiatric intervention. Cheers,
    – user37981
    Commented Oct 12, 2020 at 20:44

You ask at the end whether or not it is "appropriate" to address the subject in modern science.


  • Psychology extensively addresses why humans do things. The normal list of goals is "Describe, explain, predict, and control." The second and third directly engage with what humans desire.

  • Behavioral Economics is another example of a field that is focused on why people desire specific things and how they desire them.


  • Metaphysical theories are non falsifiable and as such don't fall under the category of science.
  • We don't yet know enough about psychology to attempt to truly engage with the fourth cause.

In modern evolutionary biology, teleology is considered to be an irrelevant explanation for what caused things to be. Nevertheless, you may find many biologists speaking in teleological terms - "The animal evolved wings so it could fly away from predators." or "Our brains became bigger to allow us to function better socially." This is a trick that evolutionary biologists use: It turns out that most language is clumsy and inconvenient for talking about the dynamics of evolution by selection, and evolution really looks almost as if there was teleology.

The key word is "almost". There are subtle differences between how an evolution with teleology and without teleology would look like. Biologists use the "trick" because they don't like being wordy and think that they are smart enough to never overlook the teleological fallacy.

Teleology is regarded as irrelevant because nature has no mechanism for deciding to obtain certain outcomes beforehand. Evolution is not a means to an end, it is a consequence of some individuals making imperfect copies of themselves more often than other individuals. Whenever one has a population of things, and the following claims hold for this population:

  • The things make copies of themselves. (reproduction)
  • Some the things reproduce more than others. (differential fitness/selection)
  • Copies can end up being is slightly better or worse at copying than the thing they were copied from. (mutation)

One will observe, after following this population, that the nature of the things making up the population changes over time. This is a bit of "duh" conclusion because, well, of course you will end up with a lot of the fast-reproducing ones and few of the slow-reproducing ones, and of course you may get a lot of much faster reproducing things than any in the initial population, because the fast-reproducing ones make more copies of themselves (even if some are a bit mutated). This also applies to theoretical constructs such as numbers as well as actual biological organisms, and everything in between (and is supported by ample empirical evidence at different levels). Copious repeated applications of this theme, adapted to individual cases, are pretty much the business of evolutionary biology, and these days, it's rare to find a biology that isn't even a little bit evolutionary biology as well.

It isn't hard to see that in the above example, nobody tried to make the fast-reproducing things dominate, and nobody wanted this outcome. It's just what happened, because of different rates of reproduction they have. If you get down to the differential systems describing the system, you will see that setting it up in this way literally implies that the number of fast-reproducers will increase over time (and this is really what it means "to reproduce rapidly").

Moreover, the outcome of an event cannot also be its cause, because the cause is often understood to precede the outcome in time. My investment of money in stocks last month cannot have been caused by the profit I made today, because last month, I hadn't yet made the profit. The cause can be my anticipation that I will profit, but again, evolution does not require nor seems to exhibit evidence of any "anticipation" - it will continue to repeatedly generate mutants that are unfit even though they are "clearly" not going to work out well, and evolution still works in spite of this. Teleological arguments may explain the actions of rational actors, but who wants to prove that nature is rational and/or that God exists without making circular references to evolution?

The question "why did this feature of this organism evolve?" comes up again and again in biology. Besides simple curiosity, it also has practical implications, such as "what if this feature stops working?" or "what features would, if they stopped working, produce this effect?" (for causes of disease).

When you are trying to to figure out the function of something by asking why that something is there (to clear away some of the philosophical issues, by function I mean how it helps and how its absence hurts), if your explanation starts with "it came to be so that it could..." then you know you are wrong, because nature cannot anticipate what will provide which benefit. It simply tries everything, and keeps the ones that work.

Therefore, to explain why something in biology exists and is the way it is, you have two options:

  1. Assert that it arose by random chance. (this is actually the case in reality often enough, it is known as genetic drift) The challenge with this option is explaining how it arose by chance despite being unlikely. If you succeed, then you can expect that removing the feature, or at least replacing it with an alternate version will have no effect (for example, it doesn't really affect your life if you are capable of rolling your tongue or not).
  2. Assert that it evolved. There are two challenges. The first is to explain how individuals which possessed the feature reproduced more rapidly than those who did not. The second is to explain the same for every intermediate form, because evolution happens in small steps - it is not enough to show that having wings is better than not having wings, you must also show that having wing stumps is better than not having them, because the wing had to have come from "somewhere". If you succeed, you can predict what problems having a defect in that feature will create, and after building a large enough database of such defects, you can then take any defect and immediately narrow down its cause to a handful of features known to lead to it. You can also suggest how to improve the feature further.

Some obvious ways in which teleological explanations can fail are:

  1. Without knowing, you may come to believe that a feature came to be to achieve a goal, when in reality it has no function ("why do men have nipples?"). You will then end up chasing after "red herrings" when looking for causes of effects.
  2. You will be puzzled why seemingly desirable features do not exist ("why don't animals have wheels?").
  3. You will assume that every feature is "as good as it gets", and then you will be surprised when some of them later manage to get better.
  4. You will accept as true some fantastical explanations which are impossible because the intermediate forms are too unfit, since you will assume that nature will "grin and bear it" through those "tough times" of very low fitness "for the sake of" the final outcome with very high fitness. (more broadly, this problem is known as "climbing a hill" in optimization)

That's not to say you don't have to look hard to find pitfalls of teleological thinking (although it's clearly illogical after considering the nature of evolution). That's why the "trick" exists - most of the time, you can carry on happily thinking teleological thoughts and never see any ill-consequences. But that one time when it causes you to be wrong is a possibility scientists do not like to accept. Being a heuristic, the pseudo-teleology trick is not very persuasive when you are trying to settle one argument and move on to the next gene already.

  • This response uses a tendentious definition of teleology. You seem to think a system is teleological if and only if it is designed for a goal or purpose by some conscious agent. None of the contemporary thinkers who endorse teleological explanations in biology think that. Rather they think that a system is teleological if and only if one has to invoke functional mechanisms to explain it. Incisor teeth don't appear to be for ripping meat-that's what they actually are for. But that doesn't mean they were DESIGNED to do that.
    – user5172
    Commented Jan 8, 2014 at 11:46

The following reflects a more Aristotelean/Thomist point of view:

Final or formal causality are outside the domain of science, meaning that science has nothing to say about them. Another way of saying the same thing is that science is not a method to uncover or explain formal or final causality, one of the reasons being that science is not about understanding what things are but more about explaining how they came to be. Note that by "what things are" refer to their ontology, not to simply the the material they are made of or their shape.

Were there not to be formal or final causality, the world would look like the world of Sartre's hero in Nausea, for whom people and things are to be perceived as just blobs of matter with meaningless shapes.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .