You are viewing a free preview of this lesson.
Subscribe to unlock all 10 lessons in this course and every other course on LearningBro.
The Design Argument — also called the Teleological Argument, from the Greek telos, meaning end, goal or purpose — reasons from the apparent order, regularity and purposive arrangement of the natural world to the existence of an intelligent designer, identified with God. It is an a posteriori argument: it begins not from the bare concept of God but from features of the world we actually observe, and infers a designer as the best explanation of them. The argument has an immediate intuitive pull — a complex, finely-adjusted object naturally invites the question "who made this?" — and that intuition is precisely what the AQA specification asks you to examine and test. For the 7062 course the named focus is Paley's analogical argument and Hume's criticisms, with Darwin and Dawkins supplying a modern naturalistic challenge. This lesson sets out Paley's argument with care, develops Hume's objections in detail, and reaches a judgement about whether the argument survives them. Throughout, the AO2 task is not merely to describe the argument but to weigh it: does the order of the world point beyond itself to a designer, or can it be accounted for without one?
William Paley (1743–1805) opens his Natural Theology (1802) with one of the most famous thought-experiments in philosophy. Crossing a heath, he writes, suppose I strike my foot against a stone and am asked how the stone came to be there. I might reasonably answer that, for all I know, it had lain there forever; the answer would not obviously be absurd. But suppose I had instead found a watch upon the ground. Here the same answer — that it had lain there forever — would not satisfy. Why the difference? Because, Paley says, when we inspect the watch we perceive that its parts are framed and put together for a purpose: the springs, the wheels, the chain, the glass face are arranged so as to produce regular motion and tell the time. Remove or alter the parts and the mechanism fails. This adaptation of means to an end is something a stone does not exhibit.
Key term: Teleological argument — an a posteriori argument that infers the existence of an intelligent designer (God) from observed order, regularity or purposive arrangement in the natural world.
Paley insists on a point often missed by careless readers: the inference to a watchmaker does not depend on our ever having seen a watch made, nor on the watch being perfect, nor on our being able to work out the function of every wheel. Even a watch that sometimes went wrong, or contained parts whose use we could not discover, would still compel us to infer contrivance. The "design" we detect is the purposive fitting together of parts, and that is enough to demand a designer.
He then turns the analogy outward. The works of nature, Paley argues, surpass the works of human art in their complexity and the subtlety of their contrivance — and they do so "in a degree which exceeds all computation." His central illustration is the eye. The eye is, in effect, an optical instrument: it focuses light by means of a lens whose curvature is adjusted, regulates the quantity of light admitted by means of the iris, and resolves images upon a light-sensitive surface. Paley compares it explicitly to the telescope, observing that the eye exhibits every mark of design that the telescope does, only more so. If the telescope had a maker, the eye — far more intricate, and self-maintaining — must have one too. From the contrivance of natural things he infers a contriver: an intelligent, powerful Being who stands to nature as the watchmaker stands to the watch.
| Feature of the inference | The watch | The natural world |
|---|---|---|
| Observation | Parts arranged to a purpose (telling time) | Parts arranged to a purpose (e.g. the eye seeing) |
| Mark of design | Adaptation of means to an end | Adaptation of means to an end, on a vaster scale |
| What it survives | Imperfection, ignorance of some parts | Imperfection, ignorance of some parts |
| Inference | An intelligent watchmaker | An intelligent designer — God |
Paley's argument is best classified as a design qua purpose argument: it fixes on the purposive adaptation of parts within an organism. It should be distinguished from a design qua regularity argument, which fixes instead on the law-like, repeated patterns of the cosmos as a whole (the orbits of the planets, the succession of the seasons). The distinction matters because Hume's objections bite hardest against the qua purpose analogy, while later thinkers such as Swinburne try to relocate the argument onto regularity.
Key term: Analogy — an argument that, because two things resemble each other in some respects, they probably resemble each other in a further respect. Paley's whole case rests on the strength of the resemblance between the natural world and a designed machine.
It helps to be explicit about the logical form of Paley's reasoning, because it determines what counts as a good objection. The argument is inductive and analogical, not deductive: it does not claim that a designer follows with logical necessity from order, only that a designer is the most reasonable explanation of it. Its strength is therefore a matter of probability and the quality of the resemblance, not of strict entailment. This has two consequences. First, the argument cannot be refuted simply by pointing out that its conclusion is not certain — no inductive argument yields certainty. Second, and more dangerously for Paley, the argument is only as good as the analogy is close: the more the natural world differs from the human artefacts we know to be designed, the weaker the inference. Paley needs the resemblance between an organism and a machine to be tight; his critics will argue that it is loose. A further subtlety is that Paley's argument has a single-case structure on the side of the conclusion — there is one universe — even though it rests on many cases on the side of the premise (we have seen many watches made). This asymmetry is precisely the wedge Hume drives in.
Notice, too, what Paley is not claiming. He is not arguing from gaps in our scientific knowledge ("we cannot explain the eye, therefore God"). His claim is the opposite: the more we understand the eye's exquisite adaptation, the more it impresses us as contrivance. This matters for evaluation, because it means the cheap charge of a "God of the gaps" does not straightforwardly apply to Paley — though, as we shall see, it does apply to later versions such as Behe's "irreducible complexity," which do rest on what natural selection has not yet explained.
Several thinkers beyond the specification's named core help locate Paley's argument in a wider tradition. Aquinas' Fifth Way (in the Summa Theologica) is an earlier qua regularity version: Aquinas observes that natural bodies which lack intelligence — such as an arrow — nonetheless act towards an end, and just as an arrow is directed by an archer, so unintelligent natural things must be directed to their ends by an intelligent being, "and this being we call God." F.R. Tennant (1866–1957), in Philosophical Theology, developed an anthropic and aesthetic version: the universe is intelligible to reason, provides for the needs of life, and contains a wealth of beauty far exceeding what survival requires — all, he argues, pointing to a designing mind. Richard Swinburne (b. 1934) restates the argument probabilistically as a temporal-order argument: that the universe behaves according to simple, consistent laws at all is, he contends, more to be expected if God exists than if it is mere brute chance, so the orderliness raises the probability of theism. These enrichments matter for evaluation because they attempt to shift the argument onto ground that Hume's machine-analogy objection does not so easily reach.
David Hume (1711–1776) wrote his Dialogues Concerning Natural Religion (published posthumously in 1779) before Paley's Natural Theology, yet his criticisms anticipate and undercut Paley's version with remarkable precision. The objections are voiced chiefly by the character Philo, and they are the heart of what AQA expects you to deploy. They should be understood not as a single objection but as a battery of distinct attacks on the analogy and on its conclusion.
1. The analogy is weak. Hume's foundational point is that the argument is only as strong as the resemblance it rests on, and the resemblance is poor. We are entitled to infer a designer for a machine because we have wide experience of machines being made and know the kind of cause they require. But the universe is not evidently like a machine; it equally resembles a growing organism — a vegetable or an animal — and we do not infer that a carrot or a fawn was deliberately designed. Crucially, the universe is a unique object: we have no experience of universes being produced, no other universes to compare it with, and so no basis for confidently inferring what sort of cause it must have. An inference from many sampled cases is reliable; an inference from a single, unrepeatable instance is not.
2. The Epicurean hypothesis: order can arise by chance. Hume revives an ancient suggestion (associated with Epicurus). Suppose matter is finite but time is effectively infinite, and the particles of matter are in constant motion, falling into combination after combination. Over an unimaginably long stretch, the particles will pass through a vast range of arrangements — including stable, ordered, self-sustaining ones. Any universe that did settle into a lasting order would, by definition, present the appearance of design to observers within it; the order requires no external designer, only time and motion. This is a genuinely rival explanation of cosmic order, not a mere quibble.
3. The designer need not be the God of theism. Even granting that the universe is designed, the argument cannot deliver the God of classical theism. By Hume's own principle that we may attribute to a cause no more than is required to produce the effect, the most we could infer is a cause proportioned to the world — and the world is finite, flawed and mixed. So we have no warrant to infer an infinite designer (a finite world needs only a finite cause), nor a perfect one, nor even a single one: just as a ship is built by many hands, the world might be the joint work of a committee of lesser deities. It might equally be the bungled first attempt of an infant deity, or the worn-out production of an aged one. The argument, Hume insists, simply does not reach monotheism, omnipotence or perfect benevolence.
4. The presence of evil. If we read the designer's character off the world (as the argument invites), the evidence is not encouraging. The world contains vast suffering, waste and apparent cruelty. A designer inferred from this world might as easily be indifferent, or limited in power or goodness, as perfectly benevolent. The problem of evil thus re-enters as a direct objection to the design argument, corroding the move from "a designer" to "a good God."
5. The designer itself would need a designer. Hume presses a regress objection that anticipates a famous modern version (Dawkins later calls it the "ultimate Boeing 747 gambit"). If order in the world requires an explanation in terms of an ordered mind, then the designing mind — which is at least as ordered and complex as the world it designs — must itself require explanation. Either we must posit a designer of the designer, and so on without end, or we must allow that some ordered thing can exist without an external designer. But if we allow the latter, why not stop at the universe and treat its order as needing no further explanation? The argument, Hume suggests, does not actually explain order; it merely relocates it into a mind and then halts arbitrarily.
6. Like effects imply like causes — a human, embodied, fallible designer. Hume's principle that we infer causes proportioned to their effects cuts further than the question of power. Our only experience of minds designing things is of human minds — embodied, finite, mortal, working by trial and error, often in teams, frequently improving on the botched efforts of predecessors. If we are to reason by strict analogy from designed artefacts to the designer of the universe, consistency demands that we attribute to the cosmic designer the features that actually accompany design in our experience: a body, limited intelligence, mortality, and the use of prior models. The very rigour of the analogy thus drags the conclusion away from the immaterial, infinite, unique God of theism and toward something embarrassingly anthropomorphic. The believer wants the analogy strong enough to deliver a designer but weak enough not to deliver a human-like one, and Hume's point is that one cannot have it both ways.
It is worth pausing on how these objections fit together, because in an exam the temptation is to list them. They form two clusters. Objections (1) and (5)–(6) attack the logic of the analogy itself: the resemblance is too weak to license the inference, the inference does not really explain order, and where it does reason rigorously it yields the wrong kind of designer. Objections (2)–(4) supply rival or deflating conclusions: order can arise without design (Epicurus), and even granting design the cause need not be one, infinite, or good. The strongest evaluation will note that objection (2) is the one Darwin later transformed from a bare possibility into a worked mechanism, and that objections (3), (4) and (6) survive even if a designer is granted — they limit what the designer can be shown to be.
| Hume's objection | Core claim | What it targets |
|---|---|---|
| Weak analogy | The universe is unique and as like an organism as a machine | The resemblance Paley's inference rests on |
| Epicurean hypothesis | Time + matter in motion can produce order by chance | The need for a designer at all |
| Designer ≠ God of theism | A finite, flawed world warrants only a proportioned cause | The leap to one, infinite, perfect God |
| Evil | The world's suffering counts against a benevolent designer | The goodness of the inferred designer |
| Regress | An ordered designer would itself need explaining | Whether the argument really explains order |
| Anthropomorphism | Strict analogy yields an embodied, fallible designer | The immaterial, perfect nature of the designer |
The deepest modern challenge comes from biology. Charles Darwin (1809–1882), in On the Origin of Species (1859), supplied a mechanism — natural selection — by which the appearance of design can be produced without a designer. Organisms vary; some variations confer a survival or reproductive advantage in a given environment; those variants leave more offspring; over very many generations the advantageous traits accumulate. The result is organisms exquisitely "fitted" to their conditions — the very fit that Paley took as the fingerprint of design. The eye, Paley's flagship example, can in principle be reached by a long graded sequence of incremental improvements, each conferring some advantage, from a light-sensitive patch to a focusing organ. The "design" is real, but it is apparent design, the product of an undirected algorithm rather than an intending mind.
Key term: Natural selection — the process by which heritable traits that improve an organism's chances of surviving and reproducing become more common in a population over generations, producing adaptation without conscious design.
Richard Dawkins (b. 1941) sharpens the point in The Blind Watchmaker (1986), a title chosen as a deliberate riposte to Paley. The processes of nature, Dawkins argues, have no foresight, no purpose and no mind; natural selection is the "blind watchmaker" — blind because it does not plan ahead, a watchmaker because it nonetheless produces the appearance of contrivance. For Dawkins, Paley was not foolish to be impressed by biological complexity; he was simply wrong about its cause, because the true explanatory mechanism was not discovered until 1859. The force of this objection is that it does not merely raise the bare possibility of order without design (as Hume's Epicurean hypothesis did); it offers a detailed, evidenced, working alternative. Where there is a successful naturalistic explanation, the inference to a designer loses its claim to be the best explanation.
It is important to be precise about the limits of this challenge. Darwinian evolution addresses biological design — the adaptation of living things. It does not directly touch a qua regularity argument about why there are stable laws of nature at all, nor the fine-tuning of cosmological constants, and so it does not by itself refute Aquinas, Tennant or Swinburne. A theist may also argue that evolution is the means God chose to design — that natural selection and divine purpose are compatible rather than rival. The atheist replies that once natural selection is doing the explanatory work, God becomes an unnecessary extra hypothesis. This is the live edge of the debate, and it is worth seeing exactly why. Darwin removes the need for a designer to explain adaptation: he shows how the appearance of foresight can be produced by a mindless, cumulative process operating on heritable variation. But Darwin presupposes the very things a regularity argument is impressed by — that there are stable laws, that like causes produce like effects, that matter has the dispositions it does. Natural selection cannot operate at all unless the universe is already law-governed. So the defender of design may grant Darwin everything at the biological level and yet ask: why is there a cosmos orderly enough for natural selection to work in the first place? The disagreement is therefore not really about evolution as biology, which both sides accept, but about whether the background order on which evolution depends itself calls for explanation.
The most discussed modern restatement of design is the fine-tuning or anthropic argument, which is the natural heir of Tennant's first kind of evidence and of Swinburne's emphasis on regularity. The claim begins from a much-remarked feature of modern cosmology: the fundamental constants and quantities of physics — the strength of gravity, the strength of the strong nuclear force, the cosmological constant, the mass-difference between the proton and neutron, the precise rate of expansion just after the Big Bang — appear to fall within extraordinarily narrow ranges compatible with the eventual emergence of stars, chemistry and life. Had several of these values been very slightly different, the universe would have recollapsed almost immediately, or consisted only of hydrogen, or contained no stable atoms at all, and there would have been no possibility of life. The universe looks, in this technical sense, "fine-tuned" for life.
Key term: Anthropic / fine-tuning argument — a modern qua regularity design argument: the fundamental constants of physics lie within the narrow ranges necessary for life, and this apparent fine-tuning is best explained by a designer who intended life.
The argument is usually framed as an inference to the best explanation among three rival hypotheses for the fine-tuning: chance (the values just happened to be right), necessity (the values could not have been otherwise — though we have no theory showing this), or design (a designer set the values to permit life). Proponents such as Swinburne argue that chance is wildly improbable, necessity is unsupported, and so design is the most reasonable explanation: a life-permitting universe is much more to be expected if there is a God who values conscious life than if the constants were set blindly. This restatement deliberately sidesteps Darwin (it is about cosmic parameters, not biological adaptation) and sidesteps Hume's machine-analogy (it does not rely on the universe being "like a machine," only on the improbability of the values).
The standard naturalistic reply is the multiverse hypothesis. Suppose ours is not the only universe but one of an enormous — perhaps infinite — ensemble of universes, each with its own randomly assigned constants. Then it is no surprise that some universe has life-permitting values, and no surprise that we find ourselves in a life-permitting one — for we could not exist to observe a universe that did not permit observers. On this view the appearance of fine-tuning is a selection effect: given enough universes with varied constants, a life-friendly one is statistically expected, and no designer is required. This is, in effect, the Epicurean hypothesis upgraded with cosmological physics, just as Darwin upgraded it with biology.
The debate then turns on which explanation is more rational, and several moves are worth weighing for AO2. (1) The theist objects that the multiverse is ad hoc and unobservable — it posits a vast unseen reality precisely to avoid a designer, and so is no more economical or testable than theism; it may even be less economical, since it multiplies entities enormously. (2) The naturalist replies that some multiverse models drop naturally out of independently motivated physics (certain versions of inflationary cosmology and string theory), so the ensemble is not invented merely to dodge God. (3) A deeper objection — pressed by some philosophers — is that the multiverse may commit an "inverse gambler's fallacy": the fact that many other universes exist does nothing to make it more likely that this particular universe is life-permitting, so the selection-effect reasoning may be subtly flawed. (4) Against all of this, the theist's own explanation is not cost-free: invoking God raises the question why God would create this universe, and the problem of evil remains, since a fine-tuned universe is also a universe of vast suffering. The honest verdict is that fine-tuning gives the design programme renewed life, but it does not settle the matter: the choice between a designer and a multiverse is a choice between two large, contestable hypotheses, and neither can be demonstrated.
| Explanation of fine-tuning | Core idea | Main difficulty |
|---|---|---|
| Chance | The constants just happened to be life-permitting | The improbability is said to be extreme |
| Necessity | The constants could not have been otherwise | No physical theory currently shows this |
| Design | A designer set the constants for life | The problem of evil; why this universe? |
| Multiverse | Many universes; we observe a life-permitting one | Unobservable / ad hoc; possible inverse-gambler's fallacy |
Even at its strongest, a crucial question hangs over every version of the design argument: what exactly does it establish? This matters because the conclusion the believer wants is not "a designer" but "God" — the single, infinite, eternal, omnipotent and perfectly good God of classical theism. Hume's third objection already pressed the gap, and it is worth developing as an evaluative theme in its own right.
The most a successful design argument seems to deliver is a designing intelligence powerful and clever enough to produce the observed order. But "powerful and clever enough" is not "omnipotent and omniscient": a finite world might be the work of a very great but finite mind. The argument does not show the designer is one (the order might be a collaborative product), nor that the designer still exists (a watchmaker may die after making the watch; the designer might have created the universe and ceased to be), nor that the designer is good (the abundance of natural evil tells against benevolence). It certainly does not establish a designer who is worthy of worship, who intervenes in history, or who loves human beings — all features central to the theistic God. At best, then, the argument supports a thin "designer," not the rich God of the Bible; the gap between the two must be bridged, if at all, by other arguments and by revelation.
Defenders make two replies. First, Swinburne's cumulative case: no single argument proves God, but the design argument is one strand in a cumulative rope — combined with the cosmological and religious-experience arguments, each raising the probability of theism, the total case is stronger than any part. The simplicity of a single, infinite God, Swinburne adds, is actually a more probable hypothesis than a committee of finite designers, because an infinite quantity is in a sense simpler than an arbitrary finite one. Second, the theist may concede that natural theology only takes us to the threshold — to a designer — and that faith and revelation complete the journey to the full character of God. The critic responds that this concession is significant: if the argument cannot by itself reach God, then "the design argument proves God exists" is, strictly, false, even if the argument has some lesser evidential value.
| Issue | Paley (and supporters) | Hume / Darwin / Dawkins |
|---|---|---|
| Source of order | Purposive contrivance demands a contriver | Chance + time (Hume) or natural selection (Darwin) |
| Strength of analogy | The world is strikingly machine-like | The world is also organism-like; and it is unique |
| What is established | An intelligent, powerful designer | At most a finite, possibly plural, designer — or none |
| Goodness of designer | A benevolent God | Evil counts against benevolence |
| Status of the inference | Best explanation of order | Not the best explanation once selection is available |
Question: "Hume's criticisms destroy the design argument." Evaluate this view.
Mid-band approach would list Paley's watch and then list Hume's objections side by side, concluding vaguely that "there are points on both sides." This describes rather than evaluates: it does not weigh whether any particular objection actually succeeds, and it reaches no reasoned judgement.
A Stronger / Top-band response argues a sustained line. We might begin by granting that Hume's objections are powerful precisely where Paley is most exposed — the qua purpose analogy. Hume's point that the universe is unique, and so an unsafe basis for inference, strikes directly at the analogical form: Paley reasons from many watches to one watchmaker, but we have exactly one universe and no comparison class, so the inductive inference is weak. Even granting design, Hume shows that the conclusion overreaches: a finite, flawed world licenses at most a finite, perhaps plural, perhaps morally indifferent designer, not the single, infinite, benevolent God of theism. On the analogical version, then, Hume does serious damage.
However, we should resist the claim that Hume "destroys" the design argument as such, for two reasons. First, Hume's sharpest objection — the Epicurean hypothesis — was decisively upgraded, not by Hume but by Darwin: the real threat to Paley is not the bare logical possibility of chance order but the detailed mechanism of natural selection, which explains biological adaptation without a designer. It is Darwin, more than Hume, who removes Paley's flagship evidence. Second, Hume's objections target the qua purpose analogy; they leave largely untouched a qua regularity argument about why there are stable, intelligible laws at all. Swinburne's probabilistic reformulation accepts evolution and relocates the argument onto temporal order, arguing that the law-governedness of the universe is more probable given theism. Whether that succeeds is contestable — the atheist may reply that "the laws are just brute" — but it shows the argument is not simply destroyed; it migrates.
The judgement, then, should be discriminating. Hume's criticisms, completed by Darwin, are fatal to Paley's version: as an inference from biological contrivance to the God of classical theism, the design argument fails. But "the design argument" is broader than Paley, and a reformulated qua regularity or fine-tuning version remains a live, if inconclusive, contender. The most defensible verdict is that Hume fatally wounds the classical analogical argument while leaving the wider design programme standing but unproven.
Examiner-style commentary: What lifts this response is that it discriminates between versions of the argument rather than treating "the design argument" as one undifferentiated thing — this is the move that separates Top-band from Mid-band. It correctly identifies that Darwin, not Hume, delivers the decisive blow to Paley (a precise, accurate piece of attribution), engages each objection on its merits, and reaches a substantiated judgement that is qualified rather than sweeping. The structure is an argument, not a list: each paragraph advances the case toward the conclusion.
Exam tip: Distinguish design qua purpose (Paley) from design qua regularity (Aquinas, Swinburne) early in any answer — it lets you show that an objection which sinks one version may leave the other afloat. Be accurate about who says what: it is Darwin, not Hume, who provides the mechanism for order-without-design, and Hume who shows the inference cannot reach the God of theism. Top-band answers reach a qualified judgement, not a blanket "it works / it fails."