Q
Question
Recognise the gap that A2's closing question opened — and understand why closing it matters

Article A2 ended with a provocation. A deductive argument can be valid — meaning its conclusion follows necessarily from its premises — and yet still deliver a false conclusion. An inductive argument can marshal impressive evidence and still turn out to be wrong. If validity and strength are not enough to make an argument good, what is?

The answer requires distinguishing two separate questions that it is easy to run together. The first question is about logical form: does the conclusion follow from the premises? The second question is about the real world: are the premises actually true? An argument is only fully trustworthy when the answer to both questions is yes — when the form is sound and the premises are genuinely true.

Consider this argument, which is perfectly valid:

A Valid Argument With a False Conclusion Deductive
P1All fish are mammals.
P2Sharks are fish.
∴ CTherefore, sharks are mammals.
The conclusion is false. But the argument is valid: if P1 and P2 were both true, the conclusion would have to be true. The problem is P1 — it is false. Fish are not mammals. The argument has good form but bad premises. Validity alone is never enough.

Now consider an argument in the other direction — one whose premises are true but whose form is broken:

True Premises, Invalid Form Deductive
P1All dogs are mammals.
P2Cats are mammals.
∴ CTherefore, cats are dogs.
Both premises are true. The conclusion is obviously false. The argument is invalid: the conclusion simply does not follow from these premises. P1 tells us that all dogs are mammals; P2 tells us cats are mammals — nothing in either premise entails that cats are dogs. True premises and false conclusion, together, expose an invalid argument.
A good argument needs two things at once: premises that are actually true, and a logical structure that genuinely connects them to the conclusion. Either alone is insufficient.

The two examples above nail down what the question of this article is really asking. Evaluating an argument requires two independent assessments — one about logical form, one about factual truth — and the vocabulary that philosophers use to make those assessments precisely is the subject of everything that follows.

U
Unpack
Define validity, soundness, strength, and cogency precisely — and map how they relate to each other

Evaluating deductive arguments: validity and soundness

For deductive arguments, the two dimensions of evaluation each have a name. The logical form question — does the conclusion follow necessarily from the premises? — is answered by the concept of validity. The truth question — are the premises actually true? — is answered by looking for soundness.

V
Validity (deductive arguments only)
A deductive argument is valid if it is impossible for all the premises to be true and the conclusion to be false at the same time. Validity is a property of the argument's logical form, not its content. An argument can be valid even if its premises are false, and even if its conclusion is false — provided that the falseness of the conclusion would have to trace back to false premises rather than to a breakdown in logical structure.
S
Soundness (deductive arguments only)
A deductive argument is sound if and only if it is (1) valid AND (2) all its premises are actually true. Soundness is the gold standard for deductive arguments: a sound argument is one you cannot rationally reject, because its conclusion follows necessarily from premises you have no grounds to deny. If an argument is sound, its conclusion must be true.

Soundness is the goal of deductive philosophical argument. When philosophers reconstruct and evaluate arguments, they are asking: is this argument sound? That single question breaks into two: is it valid? and are the premises true? Both must be answered yes for the argument to fully succeed.

This also explains why validity alone never settles a philosophical dispute. Someone who disagrees with a valid argument's conclusion does not have to accept it — they can instead challenge one of the premises. The history of philosophy is largely a history of exactly this move: examining arguments that appear valid and asking which premise is false, or insufficiently supported, or ambiguous.

The four possible combinations for deductive arguments

Because validity and truth of premises are logically independent properties, there are four possible combinations. Understanding all four is essential for philosophical argument analysis.

Valid form
Invalid form
All premises true
Sound ✓
"All humans are mortal. Socrates is human. ∴ Socrates is mortal."
The gold standard. Valid form + true premises guarantees a true conclusion. This is the only combination that fully succeeds as a deductive argument.
Invalid (true premises)
"All dogs are mammals. Cats are mammals. ∴ Cats are dogs."
True premises, false conclusion — proof of invalidity. If premises are true and the conclusion is false, the argument cannot be valid. This combination is logically impossible for valid arguments.
At least one premise false
Valid but unsound
"All fish are mammals. Sharks are fish. ∴ Sharks are mammals."
Good logical form, bad premise. The argument is valid — the conclusion would follow if P1 were true — but P1 is false, so the argument fails to establish its conclusion in the real world.
Invalid and unsound
"All philosophers are boring. Some interesting people exist. ∴ No philosophers are interesting."
Broken on both dimensions: the conclusion doesn't follow from the premises, and at least one premise is false. Doubly defective.

One further point about validity deserves emphasis: you cannot tell whether an argument is valid by checking whether its conclusion is true. A valid argument with a false premise can have a true conclusion, a false conclusion, or anything in between. Validity is a purely structural property — it asks only about the relationship between premises and conclusion, not about the conclusions of either.

Evaluating inductive arguments: strength and cogency

Inductive arguments are evaluated using parallel but distinct concepts. Because inductive arguments can never guarantee their conclusions — only make them more or less probable — the relevant question is not whether the conclusion must follow from the premises, but how well the premises support the conclusion. This is measured by strength.

Str
Strength (inductive arguments only)
An inductive argument is strong if the premises, assuming they are true, make the conclusion highly probable. Strength comes in degrees — inductive arguments are not simply strong or weak but stronger or weaker, depending on factors like sample size, sample diversity, the track record of the generalisation, and whether the observed pattern has an underlying causal explanation. Unlike validity, strength can be increased by adding more supporting evidence.
Cog
Cogency (inductive arguments only)
An inductive argument is cogent if it is (1) strong AND (2) all its premises are actually true. Cogency is the inductive equivalent of soundness — the gold standard for inductive arguments. A cogent argument is one whose premises are both true and sufficient to make the conclusion highly probable. It is the best an inductive argument can be, and it still cannot guarantee its conclusion.
The Spectrum of Inductive Strength
Very weak Weak Moderate Strong Very strong
Very weak
"I met one rude Australian once. Therefore Australians are rude people." — Sample of one, wildly unrepresentative.
Weak
"In my experience, philosophy students tend to enjoy reading. Therefore most philosophy students probably enjoy reading." — Small, personally biased sample.
Moderate
"In a survey of 500 university students, 68% reported that they found critical thinking skills useful in their careers. Therefore critical thinking is probably useful for most graduates." — Reasonable sample, some selection bias.
Strong
"In double-blind clinical trials across twelve countries involving 8,000 patients, aspirin reduced fever in 91% of cases. Therefore aspirin will probably reduce this patient's fever." — Large, diverse, controlled sample.
Very strong (approaching near-certainty)
"Every physical object ever studied has been subject to gravitational attraction. Therefore the next physical object studied will almost certainly be subject to gravitational attraction." — Vast, diverse evidence base; confirmed mechanism; no counter-examples in centuries of investigation.
E
Examine
Apply the evaluative framework to real philosophical arguments — including one of the most discussed in the Western tradition

Abstract concepts only become useful when you can apply them to real arguments. The following three worked evaluations progress from simple to genuinely contested — the last is one of the most debated arguments in the entire history of Western philosophy.

Evaluation 1: A straightforward case

The Socrates Argument — revisited from A1 Deductive
Argument in standard form
P1All humans are mortal.
P2Socrates is human.
∴ CTherefore, Socrates is mortal.
1
Is it valid? (Can we accept all premises and deny the conclusion?)
No — we cannot. If all humans are mortal and Socrates is human, Socrates must be mortal. There is no possible world in which both premises are true and the conclusion is false. The argument is valid.
2
Are all the premises true?
P1: Every human being who has ever lived has died. No human has been observed to be immortal. True. P2: Socrates was a human being. True. Both premises are true.
Verdict Sound ✓ Valid form + true premises = the conclusion (Socrates is mortal) is established with certainty. This is what a successful deductive argument looks like.

Evaluation 2: Aquinas's Cosmological Argument

Thomas Aquinas, the thirteenth-century philosopher-theologian, offered five arguments for God's existence in his Summa Theologica. The most philosophically significant is the First Way — the argument from motion (or more precisely, from change). It is a model of careful medieval logical thinking, and has been debated continuously for seven centuries.

TA
Philosopher
Thomas Aquinas
1225–1274
Natural theology  ·  Metaphysics  ·  Ethics  ·  Aristotelian synthesis
Aquinas attempted to demonstrate the existence of God through purely rational argument — without appealing to faith, scripture, or authority. His Five Ways (Quinque viae) in the Summa Theologica (1265–1274) are among the most carefully constructed philosophical arguments in the Western tradition, drawing heavily on Aristotle's metaphysics. Whether they succeed is one of the enduring questions of philosophy of religion. The First Way — the argument from motion — is the simplest and most purely logical of the five, and the best entry point for argument analysis.
Relevant work: Summa Theologica, Part I, Question 2, Article 3 — "Whether God Exists." Aquinas presents each argument in compressed, structured scholastic form: Objection → Reply → Main argument. The Five Ways occupy a single page and reward very close reading.
Aquinas's First Way — The Argument from Motion Deductive
Argument reconstructed in standard form
P1Everything that is in motion (changing) is put in motion by something else — nothing moves itself.
P2An infinite regress of movers is impossible — there cannot be an endless chain of things each moved by the next with no first mover.
P3The world contains things that are in motion.
∴ CTherefore, there must be a First Mover — something that moves without itself being moved. This is what everyone calls God.
1
Is it valid?
Largely yes. If we accept all three premises, we are logically committed to the existence of something that initiates motion without itself being moved. The logical structure is solid: the premises, if true, appear to force the conclusion. Most philosophers who contest this argument attack the premises, not the logical form.
2
Is P1 true? (Nothing moves itself)
Contested. Aquinas is drawing on Aristotle's concept of "motion" as change in general — not merely physical movement, but any transition from potentiality to actuality. In modern physics, particles can spontaneously change state (quantum vacuum fluctuations, radioactive decay) without an external cause. Whether these phenomena constitute genuine counter-examples to P1, or whether "motion" in Aquinas's metaphysical sense is different enough from physical change to survive, is an active philosophical debate.
3
Is P2 true? (Infinite regress is impossible)
Deeply contested. Aquinas claims an infinite causal regress is metaphysically impossible — that there must be a starting point. But why? In mathematics, infinite series are not only possible but essential. Many philosophers (including Hume and Bertrand Russell) have argued that Aquinas simply asserts rather than proves the impossibility of infinite regress. If infinite regress is coherent, the argument for a First Mover collapses. This is the premise most philosophers find weakest.
4
Even if the argument succeeds — does the conclusion establish "God"?
A further challenge, noted even by sympathetic readers: the argument, if sound, establishes only that there is a First Mover — an uncaused cause. It does not by itself establish that this First Mover is personal, omnipotent, omniscient, benevolent, or identical with the God of any particular religion. Aquinas's final phrase — "this is what everyone calls God" — is philosophically the weakest part of the argument.
Verdict Valid but contested soundness The argument has good logical form — the conclusion follows if the premises are granted. But P2 in particular is far from obviously true, and P1 faces challenges from modern physics. Most philosophers conclude the argument is valid but not demonstrably sound. Whether that is a fatal objection depends on one's broader epistemological commitments.

Notice what the evaluation framework has done here. It has moved us from a vague sense that "there must be something wrong with the argument" to a precise identification of exactly which premise is doing the heaviest lifting and why it is vulnerable. That is the practical value of these concepts: they turn vague disagreement into precise, productive philosophical engagement.

Evaluation 3: An inductive argument from ethics

Peter Singer's Argument from Poverty — a simplified version Inductive
Argument in standard form
P1Suffering and death from lack of food, shelter, and medical care are bad.
P2If it is in our power to prevent something bad from happening, without sacrificing anything of comparable moral importance, we ought to do it.
P3It is in the power of affluent individuals to prevent deaths from extreme poverty by donating to effective charities.
∴ CTherefore, affluent individuals ought to donate substantially to effective charities.
1
Is this deductive or inductive?
This argument is best understood as deductive in its logical form — Singers intends the conclusion to follow necessarily from the premises. We evaluate it for validity and soundness. (Note: some presentations of Singer's argument are explicitly inductive, but this standard form is structured deductively.)
2
Is it valid?
Yes. If we accept all three premises, we cannot coherently deny the conclusion. P1 says extreme poverty outcomes are bad; P2 says we ought to prevent bad things we can prevent at no comparable cost; P3 says we can prevent those outcomes. Together they entail we ought to act.
3
Are the premises true? Which is most contested?
P1 and P3 are relatively uncontroversial — virtually everyone agrees poverty suffering is bad, and evidence confirms wealthy individuals can prevent deaths through targeted donations. The entire philosophical battle is over P2. Does the principle that "we ought to prevent bad things we can prevent at no comparable cost" commit us to radically demanding levels of charitable giving? Singer himself accepts this: he argues it entails giving until you reach the point of marginal utility. Critics (including Susan Wolf and Frances Kamm) argue P2 is either false or needs significant qualification — that there is a morally relevant distinction between preventing harm and merely failing to provide a benefit.
Verdict Valid — soundness of P2 is the philosophical question The argument is valid, and P1 and P3 are broadly accepted. Whether it is sound depends entirely on whether P2 is true — which is the central debate in applied ethics about the demands of morality. This is philosophy working as it should: the evaluative framework has isolated exactly the right premise to argue about.
S
Synthesise
Assemble the complete argument evaluation method from A1, A2, and A3 into a single, usable framework

You now have all the pieces. Articles A1, A2, and A3 together have given you a complete method for analysing any argument you encounter in philosophy — or anywhere else. This Synthesise stage assembles those pieces into a single, coherent framework you can apply immediately and carry through every subsequent package on this site.

The complete argument evaluation framework

The method has six steps, applied in sequence. The first two come from A1, steps three and four from A2, and steps five and six from this article.

Complete Argument Evaluation — Six Steps
1
Reconstruct the argument in standard form (A1)
Identify the conclusion first (look for conclusion indicators: therefore, thus, hence). Then identify the stated premises (look for premise indicators: because, since, given that). Make hidden premises explicit. Write it out as P1, P2 … ∴ C.
2
Interpret charitably (A1)
Give the argument its strongest possible reading before evaluating it. Do not attack a weak version of the argument when a stronger version is available. This is the principle of charity — required for honest philosophical engagement and expected in all philosophy examinations.
3
Classify: deductive or inductive? (A2)
Ask whether the premises are supposed to guarantee the conclusion (deductive) or merely support it as probable (inductive). This determines which evaluation concepts apply in steps 4–6.
4
Evaluate the logical form (A3 — this article)
For deductive arguments: is it valid? Could all premises be true and the conclusion still false? If yes, it is invalid — identify which step in the reasoning breaks down. For inductive arguments: is it strong? Do the premises, if true, make the conclusion highly probable? If not, why — is the sample too small? Biased? Is there a confounding factor?
5
Evaluate the premises (A3 — this article)
For each premise: is it true? Is it supported? Is it contested? Identify which premise, if false, would most damage the argument. This is often where the real philosophical action is — the deepest and most important objections to famous arguments almost always target a specific premise, not the logical form.
6
Deliver a precise verdict (A3 — this article)
For deductive arguments: state whether the argument is sound (valid + true premises), valid but unsound (valid form, but a premise is false or unestablished), or invalid (the conclusion does not follow). For inductive arguments: state whether it is cogent (strong + true premises), strong but uncogent, or weak. In either case, explain precisely why.

This six-step method is what every philosophy examiner is testing when they ask you to "analyse and evaluate a philosophical argument." They are asking whether you can reconstruct it faithfully, classify it correctly, test its logical form, interrogate its premises, and deliver a precise verdict with reasons. The six steps give you the vocabulary and the structure to do all of that.

One more observation worth making explicit: in philosophy, demonstrating that an argument is invalid or unsound does not by itself settle the question of whether the conclusion is true. An argument for God's existence might be unsound — and God might still exist. An argument for the permissibility of euthanasia might be invalid — and euthanasia might still be permissible. The failure of an argument establishes only that this particular route to the conclusion fails. There may be other routes. Philosophy is rarely finished by showing one argument wrong.

T
Transfer
See how validity and soundness operate outside philosophy — and carry the complete A1–A3 toolkit into the rest of this package

Validity and soundness in law

Legal reasoning is one of the closest real-world parallels to formal philosophical argument analysis. A barrister constructing a legal argument faces exactly the same two-dimensional challenge: they need both valid logical form and true (or established) premises.

In a criminal trial, the prosecution's argument has a deductive skeleton. The major premise is a rule of law: "Anyone who performs act X with intention Y is guilty of crime Z." The minor premise is a factual claim about the defendant: "This defendant performed act X with intention Y." The conclusion follows necessarily: "Therefore this defendant is guilty of crime Z." The legal system separates these two questions explicitly: the judge determines the law (the major premise — validity and its meaning) while the jury determines the facts (whether the minor premise is established — soundness).

When a defence barrister challenges a prosecution argument, they are making either a validity objection ("Even if the facts are as you say, the legal rule does not apply in this way") or a soundness objection ("The factual premise has not been established beyond reasonable doubt"). These are the same two moves as in philosophical argument evaluation — just in a different institutional context.

The complete toolkit — a summary across A1, A2, A3

After three articles, you have assembled a complete vocabulary for philosophical argument analysis. Here is the whole system in one place.

From A1: An argument is a set of propositions where premises are offered as reasons for a conclusion. Arguments can be reconstructed in standard form (P1, P2 … ∴ C). Hidden premises (enthymemes) should be made explicit. Arguments are identified by indicator words.

From A2: Deductive arguments aim to guarantee their conclusion; inductive arguments aim to make it probable. Deduction reasons from general to specific; induction from specific to general. Hume showed induction cannot be rationally justified in any non-circular way. One counter-example can destroy a universal inductive generalisation.

From A3: Deductive arguments are evaluated for validity (does the conclusion follow?) and soundness (valid + true premises). Inductive arguments are evaluated for strength (does the evidence make the conclusion probable?) and cogency (strong + true premises). Soundness and cogency are the respective gold standards.

Connecting forward to Article A4

You now know what a good argument looks like. Article A4 introduces the systematic catalogue of ways arguments go wrong — the informal fallacies. A fallacy is a form of argument that appears persuasive but is logically defective: either invalid (the conclusion does not follow from the premises) or relying on premises that are false, irrelevant, or missing.

Understanding fallacies is made much easier by having the evaluative framework of A1–A3 in hand. When you see an ad hominem fallacy — attacking the person rather than their argument — you now know precisely what has gone wrong: the premise about the person's character is irrelevant to the conclusion about whether their argument is sound. When you see a straw man — misrepresenting an opponent's argument to make it easier to attack — you know that the argument being refuted is not the one that was offered, so even a sound refutation of the straw version leaves the original untouched. The fallacies are not a random list of errors to memorise; they are applications of the evaluative framework you have just built.

The question to carry with you into Article A4
You can now tell when an argument is valid, and when it is sound. But you have seen throughout this article that many arguments that appear compelling are neither — that they succeed through misdirection, emotional pressure, or subtle equivocation rather than genuine logical force. What are the most common patterns of defective reasoning? Do they have names?
They do. Article A4 maps the most important informal fallacies — the recurring errors in reasoning that philosophers, lawyers, scientists, and critical thinkers have been identifying and naming for two and a half thousand years.