We can arrive at an infinite regress of designers as follows:

  • Suppose that X is so complex that it's considered to show evidence of design. Accordingly, we infer that an intelligent designer must be behind X. Let ID1 be this intelligent designer.
  • We can ask the following question about ID1: Is ID1 at least as complex as X? We have two possibilities:
    • if Complexity(ID1) >= Complexity(X), we can infer that there must be an intelligent designer ID2 who designed ID1.
    • if Complexity(ID1) < Complexity(X), we are admitting that complexity can increase over time, which would sound great to the ears of evolutionists, so let's just ignore this case for the purposes of this analysis.
  • So assuming the first case, we can infer that ID2 must exist. But then we can ask the same question about ID2, and infer the existence of ID3, and then the existence of ID4, and the existence of ID5, and ...
  • And thus we arrive at an infinite regress of intelligent designers.

Would ID proponents agree with this reasoning, and if so, do they find an infinite regress of intelligent designers problematic?

If one believes that infinite regresses of this kind are impossible, a possible solution would be to stop the regress by claiming that there has to be a first undesigned designer (just like the unmoved mover or uncaused cause of the cosmological arguments for God's existence), but I think this would essentially reduce ID to Creationism. Is there a way to solve the infinite regress problem without resorting to Creationism, especially in light of articles such as Intelligent Design and Creationism Just Aren’t the Same?

Just a brief update: I just came across a very similar question, with answers that provide a great complement to the answers posted here. Feel free to check it out: Dawkins on God: What are the strongest counters to his argument?


7 Answers 7


Your question contains a couple of misconceptions about ID. First, it assumes that ID is an argument for God. This is not the case. ID can certainly be used in an argument for God, and most (but not all) ID proponents are theists, but the writings on ID itself seldom try to prove the existence of God.

Second, the question assumes that the ID argument is based on the amount of complexity in a system, but this is not the case. The basic argument of ID is not that life has too much complexity (whatever that means) to have resulted from accidental causes; the argument is that life has the type of complexity that is irreducible. What this means is that there is no set of simple steps that could produce that type of complexity.

Of course Darwinism is a theory which specifically claims the opposite. That is, Darwinism contains or depends on the claim that every living organism is such that it was physically possible for that organism to evolve by a set of simple steps where each step has a reasonable probability of occurring. This is why ID proponents spend so much time criticizing Darwinism. To objectively evaluate ID, you have to read their arguments for why Darwinism fails to provide an alternative.

So in answer to your question, if the designer ID1 is irreducibly complex then, yes, there would have to be an ID2 to design ID1. However, there is no reason to suppose that ID1 has to be irreducibly complex in order to design an irreducibly complex object. There could very well be a planet where intelligent life evolved in a different direction, one in which small steps are possible in a way that they are not (claim ID proponents) on Earth. The biological arguments of ID are about the irreducibility of Earth life, not about the irreducibility of all possible life. So the specter of a vicious circle does not arise.


if Complexity(ID1) < Complexity(X), we are admitting that complexity can increase over time, which would sound great to the ears of evolutionists, so let's just ignore this case for the purposes of this analysis.

Would ID proponents agree with this reasoning, and if so, do they find an infinite regress of intelligent designers problematic?

Yes, the reasoning is sound. A secular ID proponent (a rare breed) would not find this problematic, as they would not ignore the mentioned case, and would allow for evolution. A religious ID proponent would privately welcome the infinite regress to point out that religious beliefs, like Creationism, explain away the infinite regress, with a god that is not bound by logic. In public, ID proponents might refer to Panspermia as an alternative, to distance ID as far as possible from Creationism, as that is the key to get ID and Creationism into schools science curriculum (and the ID handbook says that this narrative must be declared as "ad hominem").

In general, there is no reason to assume a designer cannot design something more complex than themselves, so a designed thing can have a designer that is less complex. Trivially a designer can design another designer by making a design that is a copy of themselves, and then adding a complexity. An example would be humans injecting plant DNA into an embryo to create a human with green skin to absorb sunlight, or even to inject computer-generated DNA into an embryo to create entirely new proteins which can boost our immune system or something.

Another obvious way a designer can design things to any complexity is cooperation. If there are 1000 designers (humans), each designing a subsystem less complex than themselves, when put together, the whole can be more complex than a single designer.

While ID aims to explain life on earth, the "logical" aspects of it are not bound to earth and humans, and thus imagining other kinds of designers which look nothing like humans and which can design something more complex than themselves needs to be contemplated.

The question is mostly about whether there can be a "minimal designer" which can design, but which cannot exist unless designed by a more complex designer. But that's all to philosophic for ID.

ID is based on the "Argument from incredulity" fallacy, following "intuition" about what evolutionary steps seem "probable", and what kind of designers they can imagine.

Is there a way to solve the infinite regress problem without resorting to Creationism, especially in light of articles such as Intelligent Design and Creationism Just Aren’t the Same?

Such articles are written for the political purpose of pushing Intelligent Design (and creationism) into school science curriculum while trying to avoid to appear like religion. See https://en.m.wikipedia.org/wiki/Teach_the_Controversy for details.

The original question already allowed for a way to resolve this without resorting to creationism. It's weird to ask a question while also providing the answer.

  • 5
    Suggest you remove this sentence: "A religious ID proponent would welcome the infinite regress to point out that religious beliefs, like Creationism, explain away the infinite regress, with a god that is not bound by logic." as it is entirely speculative and is certainly not true of any theistic ID proponent I know of. May 25, 2023 at 14:30
  • 6
    Suggest you remove this sentence: "Such articles are written for the political purpose of pushing Intelligent Design (and creationism) into school science curriculum while trying to avoid to appear like religion." as it is ad hominem, tendentious, and unjustified. May 25, 2023 at 14:34
  • 5
    @DavidGudeman The mission statement of the Discovery Institute (where that article lives) is to advance Judeo-Christian culture. And then Wikipedia does a nice job tracing ID's origins, including ""Intelligent design" was the most prominent of around fifteen new terms [Of Peoples and Pandas] introduced as a new lexicon of creationist terminology to oppose evolution without using religious language." May 26, 2023 at 2:24
  • 3
    @DavidGudeman: The historic origin of ID is well documented and supports my statements. If need be I could ad the references, but they are obvious and I want to keep the answer short.
    – tkruse
    May 26, 2023 at 4:11
  • 4
    @DavidGudeman Disallowing any commentary on motives or biases involved in arguments (by calling that ad hominems) makes it that much more difficult to change anyone's mind. Even if we can agree on nothing else, we should at least be able to agree that convincing people through logical argument alone rarely works, and many arguments are not all that logical (although we'd obviously disagree about which arguments are illogical and who can't be convinced). So there are non-logical factors involved in argumentation, that are important to talk about, for changing minds.
    – NotThatGuy
    May 26, 2023 at 12:50


I propose a thought experiment to prove my answer;

Imagine humans develop advanced, self-aware, AI systems that include robot machines. Such entities could maintain their own hardware and power supplies without human intervention. They would then live indefinitely.

Then, a biological catastrophe wipes out humanity.

As the AI robots consume the environment for their own development, all trace of the humans is erased.

We now have an obviously artificial race of beings in control of the planet.

Since we know the history of our story, we know that they arose from earlier, naturally evolved creatures.

So we have artificial creatures derived from natural creatures, with no regression.

  • 4
    That presupposes that humans are not so prohibitively complex that they couldn't have evolved, but I see your point. You can stop the regress if you are open to evolution and assume that there was a first undesigned evolved designer.
    – Mark
    May 25, 2023 at 12:47
  • @Mark Yes; since it's a thought experiment, we can just assume the precursor species is naturally evolved. To refute my argument, you'd have to prove that any species capable of creating an artificial race could not itself evolve naturally. May 26, 2023 at 13:27
  • 2
    This is a really fun thought experiment. Credibly sapient robots could also examine their own processors and conclude that, although plants and microbes appear to have evolved naturally, there are features of of the CPU which mean they're unlikely to have done the same. Sapient characters in a virtual world (the Sims version 3000) might infer the same. Importantly neither claim says anything about where the designer came from or whether they're even part of the relevant universe: they'd just claim that an intentional creative will is the best explanation for the observation that they exist.
    – Josiah
    May 26, 2023 at 22:11

Disclaimer of Bias

My bias, to be clear, is in favor of rejecting ID, but I do acknowledge that there's prima facie grounds for accepting the argument that ID is a scientific theory. ID, however, relies on certain metaphysical presumptions that fall apart under scrutiny, and when properly met, falls outside of contemporary demarcation of science (though to be accurate, if you use a 19th century definition of science, ID works just fine). You're not looking for why philosophers of science reject ID as pseudoscience, so let's answer your question instead.

Intelligent Design, Infinite Regress, and Aliens

Intelligent design naturally leads to the question, if a naturalistic and evolutionary approach to the explanation to the existence of biological organisms is insufficient to explain the existence of said organisms, where does the intelligent designer come from? So you are right in that it mimics the form of the question of the first cause. Technically, there's nothing inherently wrong metaphysically accepting infinite regress when isolated from broader questions of cosmogony. Why can't there be an infinite cycle of IDers? I don't see on the face why there can't if the question is taken without considering the position of a naturalistic universe itself. In this way, one only metaphysically rejects an infinite cycle of IDers, if one considers the broader program of materialism and naturalism in describing the universe.

But even if one accepts ID as scientific, since there's nothing philosophically objectionable about an intelligent designer on the face, and ID certainly doesn't necessarily imply creationism. One could read it as a theory that advocates that space aliens seeded our DNA (Forbes) instead. So, it's possible to argue the universe has always existed with intelligence, and that one intelligent species designs the next, ad infinitum. In my mind it begs questions, but certainly it's not an irrational first principle if one accepts the universe has this infinite property. This is actually a foundationalist argument insofar as, there is a first principle, and it sits nicely along the principle of sufficient reason. One can even kick the question of the origin of the universe down the road with quantum physical speculation about a multiverse. Our universe was caused by another universe.

Now, you're asking if this metaphysical speculation inherently disqualifies ID, and from a rational standpoint, no, it doesn't. There is no logical contradiction that forces us to eliminate infinite regress as a metaphysical necessity (SEP), and anyone who says so isn't familiar with the Agrippan arguments or deeper epistemological conversation about foundationalism, infinite regress, and circularity.

Where the Metaphysical Problem with Infinite Design Actually Lies

ID has a Medieval conception of biological category and suffers from category mistakes. That is, it presumes natural kinds (SEP) which has two ideas that are opposed by modern empirical findings. If might be understood defending:

  1. There is a privileged notion of an organism, such as the chicken, which exists independently of interpretation.
  2. It conflates chicken-as-species with chicken-as-organism, a violation of the type-token distinction.

Let's say I design a classic Roman arch to be be built of bricks with the last piece to be placed the keystone. Certainly, if I come across an arch, I can say, it's not possible for a keystone to have come to this position naturally, for if I remove it, the arch will collapse, the two sides falling apart. Thus, with arches there is "irreducible complexity" because there are no natural steps that lead to a keystone existing in such a state in nature without a designer. One can then point to a natural arch and argue see, and if I pull the section out were the keystone would be in an arch, it too falls apart, so there must be a designer just like our Roman arch was designed. There is no step in construction that prevents the top portion of the arch to get in place naturally.

The error in this analogy to the reasoning of ID is obvious. Natural arches aren't constructed and they have no keystones. This is a serious category error, and all the meaning distraction of "complexity" is a smokescreen that hides the flaw. The flaw in the reasoning has to do with categories, not infinite regression. How so?

The chicken begets the egg begets the chicken...

Infinite regress right? How can there be chickens or eggs without a supreme being creating the infinite regress in the first place? See, the infinite regress isn't logically problematic. Chickens do beget chicken eggs. And chicken eggs do lead to chickens. There's nothing irrational about that regress. It makes sense. The problem is with the categories. Simply put, a chicken-as-organism is an instance of a chicken-as-species, and so to reason about the origin of chicken-as-organism presuming that it's not a graded member of a fuzzy category chicken-as-species is the category mistake. Ultimately, by seeing that the definition of chicken-as-organism is interrelated with the genetic and statistical membership to the chicken-as-species, and that the definition of chicken-as-species arises from dinosaur-as-species to bird-as-species through evolution of dinosaurs-as-organisms towards birds-as organisms. The problem dissolves.

This is a fallacy of composition where properties of the systems (instances of times such that we consider an organism at time x) are different from the system as a whole (all instances when considering organisms and collections of organisms). To point to the heart, and say, there's no step this chicken can exist without a heart, presumes both the category of the chicken and category of the heart, and PRESUME those categories aren't fuzzy and malleable. But they are.

What is a chicken and what is a heart may be fixed in this point or range in time, but not over the entire spectrum of time. Circulatory systems start off WITHOUT hearts and gradually evolve to have them, just as dinosaurs slowly change into birds WITHOUT a privileged point in time where chickens spring into being.

Thus, the belief in natural kinds displays an ignorance of categories in the same way that intelligent design does. The problem with intelligent design is not the infinite regress, it's the category mistake it presupposes by having a static and crisp hierarchy of categories instead of presuming a continuous and fuzzy evolution of categories. That's why today in biology, cladistics is the authoritative system for understanding our the taxonomy of living beings.

From a metaphysical standpoint, there's nothing wrong with intelligent design logically prima facie if one doesn't understand fuzzy categories and presumes natural kinds, but modern philosophers have shown that such presumptions are less than adequate for contemporary scientific enterprise which overstate objective reality. 'Chickens' and 'hearts' may have objective constituency in terms of matter, but the explanatory criteria to determine their membership is conventional and utilitarian.

To show how our categories are constructs of the mind, not independently existing facts about the universe, consider how a planet like Pluto is demoted to a TNO. People who understand the nature of categories (they're nominalistic (SEP) conventions), don't object. People who struggle with more rational, less intuitive notions of categorization reject that. "Pluto was a planet when I grew up, so how can it not be a planet now?" Easy, a planet is a CATEGORY created by CONVENTION, and conventions change. That's not denying that the physical oblate spheroid doesn't exist or has changed, but the rules of the language game have because 'what is a planet' has gone from everybody gets a vote to the members of the IAU get a vote. (This also highlights the difference between folk categorization and folklore and technical categorization by domain-specific experts and jargon.)


Would ID proponents agree with this reasoning, and if so, do they find an infinite regress of intelligent designers problematic?

That seems to be a request for speculation. In any given case, I imagine the answers to those questions would be either (yes, no) or (no, moot), because (yes, yes) would be irrational.

I don't think the (yes, no) case bears much further discussion, and I'm not much interested in irrational positions, so let's talk about reasons why ID adherents might disagree with your reasoning.

Large among those is that you have oversimplified typical ID arguments. Complexity alone is not the only consideration. Arguments against the complexity of living organisms being explainable by fundamentally random, impersonal processes are generally based in part on the amount of time those random processes had to produce the specific observable results that are our evidence, and in part on the nature of our universe as we understand it.

For example,

  • An ID proponent who postulates that the entire universe was designed and created by an intelligent designer thereby places the designer outside the universe. Our observations about the universe may allow us to make inferences about the existence and nature of such a designer, but we have no context for evaluating the applicability of the ID argument to that designer. Time might not even apply to it, and there's no reason to suppose that it is governed by physical laws that govern us.

  • An ID proponent who argues that some parts of our observable universe (e.g. Earth life) must have been designed by an intelligence inside our universe and subject to its laws does not need to assume that the complexity of that intelligence is of a kind that is subject to the same ID argument. Perhaps there was enough time since the beginning of the universe for that particular being -- of a fundamentally different nature than ourselves -- to arise through random processes.

  • Along the similar lines, an ID proponent might argue that the intelligent designer could be the universe itself. That's more complex than us because it encompasses us. Although this does not speak directly to whether the universe had its own designer, those arguing against ID cannot accept the universe having been designed, as that is the very antithesis of their position. Many people do accept the evolution of the universe overall to have been fundamentally undirected, and if we accept that view together with the hypothesis that the universe is an intelligent actor, then ID does not require infinite regression.


Intelligent Design (ID) doesn't necessarily entail an infinite regress of designers, but if you want to avoid this infinite regress, any ID argument would need to address why the designer didn't need to itself be designed (otherwise this would be committing a special pleading fallacy).

People have made various attempts to make their proposed designer the terminating point by e.g. claiming that said designer is timeless or something like "properly simple" (although what exactly that means is only ever left up to one's imagination, and evidence for these claims is typically severely lacking).

If you argue that "this is too complex, it must've been designed", then you indeed run into the problem you described, where the designer is either more complex (in which case it requires a designer as well) or less complex (in which case simplicity gives rise to complexity, which wouldn't invalidate the original argument, but would undermine it quite a bit).

Irreducible complexity is another popular argument or variant of the above, and I have heard people point to various organs (e.g. the human eye) as examples of irreducible complexity. I have then seen various people explain how evolution could've or did do that (like this explanation of the evolution of the eye).

This trivially suffers from the same special pleading problem, as the designer is either irreducibly complex (requiring a designer) or reducibly complex (which raises the question of why their complexity is reducible, but the complexity of their design is not).

One could argue that design is a form of reduction in complexity, so an argument of irreducible complexity would need to be either (a) an argument specifically against an accepted theory for reducing some form of complexity, e.g. evolution (but disproving our current understanding of evolution, however colossal of a task that is, wouldn't prove design, as you'd also need to prove that they're mutually exclusive for this to be sound reasoning), or (b) an argument for design specifically, which either includes or excludes evolution (and this may be a lot more difficult than "simply" disproving evolution). In both cases, there's also a significant risk of simply saying: we don't know how it can be reduced, therefore it cannot be reduced, which would be fallacious.


There are three points to avoid infinite regress.

  1. Come upon a thing that isn't irreducibly complex that produces things that are irreducibly complex. The robot is irreducibly complex and this is not disputed. On the other hand, man makes robots and whether or not man is irreducibly complex is the subject of ongoing debate to the point where those who would argue irreducible complexity prefer other cutpoints*.

  2. We can specify an ultimate source that of its nature has no beginning. Everything has a cause is a common mistake. The correct axiom is everything that begins has a cause†.

  3. Infinite regress is not a problem if the prior history of the universe is infinite. (Spoiler: current cosmology says it is finite.) I am certain someone is going to come by and argue that an infinite past cannot be; however if you presuppose that it is, then infinite regress is the expected result.

The whole rest of this is not part of the answer but a huge set of notes. To avoid excess argument in the comments I have edited it down to bare summaries sufficient to find the material and arguments in question by content search.

*There are two favored arguments for jump-step for man above the apes: that of sudden onset language development and that of wielding fire. The thing about wielding fire is actually used in archaeology to determine if they are looking at man's encampment or apes' encampment.

However these are seldom used and much preferred are the cutpoints around biogenesis and around the laws of physics. Biogenesis requires a certain minimum complexity for a replicator and the estimates of that complexity range from high to very high. In addition, abiotic origin of amino acids has completely ignored the chirality problem. The other one is in the tuning of the constants of the universe. We are not at the point where we can write down and test via simulation alternate laws of physics but we can test the the constants and simulation results show an extremely high set of tuning of the constants. If the coupling constants are tweaked by as little as one percent in most cases the result is no metallic (in the astronomical sense) chemistry or even no atoms.

†There once was a cosmological theory that injected a mass-creating repulsive force into the laws of physics. It created new mass at the approximate rate of 1 hydrogen atom's worth/liter/billion years. The theory was valid, fair, and testable. Its current status is does not match observations and so discarded.


You must log in to answer this question.

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