Key takeaways
- A deductive argument claims that its conclusion cannot be false while its premises are true. That claim is about the form of the argument, not about whether anything in it is actually the case.
- Validity and truth are independent. A valid argument can carry a false conclusion when a premise is false, and an invalid argument can land on a true conclusion by luck.
- Soundness is the pair: a sound argument is valid and has true premises, which is the only combination that makes its conclusion safe to rely on.
- To show a form invalid, build another argument of the same shape with obviously true premises and an obviously false conclusion. One counterexample settles it.
- A meta-analysis of belief bias found that believable conclusions are endorsed more often than unbelievable ones with identical structure, and attributed the effect to a shift in response bias rather than a change in people's ability to discriminate valid from invalid.
- Our own app, IQ Test & Brain Training Games, includes logic games built on deduction and conditional reasoning, but this subject is here for the reader: the validity and counterexample checks above work on any argument, with or without an app.
A deductive argument makes one specific promise: if its premises are true, its conclusion cannot be false. That is the whole of it. The promise is about structure, not subject matter — which is why an argument can be perfectly valid and still reach a conclusion that is plainly untrue, and why an argument can be a mess and still land on something true by accident. Separating those two things is the single idea that makes deduction usable.
The Stanford Encyclopedia of Philosophy's entry on logical consequence puts the standard analysis in two parts. Necessity: the truth of the premises is necessarily sufficient for the truth of the conclusion, so it is impossible for the premises to hold while the conclusion fails. Formality: whether that holds depends on the argument's structure rather than on what its terms happen to mean.
One note before the examples, since this hub sits alongside an app. IQ Test & Brain Training Games, our own app, includes logic games with analogies, sequences, conditional reasoning and deduction, and an IQ-style test whose areas include logic; its Google Play listing calls the test's results estimates based on app activities, not clinical or professional assessments. This article is here because readers of the logic and pattern puzzles hub need the concept, and nothing in it depends on the app.
Validity is a promise about form
Consider:
Every square is a rectangle. Every rectangle has four right angles. Therefore every square has four right angles.
Both premises are true and the conclusion follows, so this is both valid and true. Now keep the shape and change the content:
Every bird can fly. A penguin is a bird. Therefore a penguin can fly.
The structure is identical and the reasoning is impeccable. The conclusion is false because the first premise is false. Validity did its job; it simply never promised anything about premises it was handed.
This is worth labouring because it is counter-intuitive. Reading the penguin argument, most people want to call it a bad argument. It is a valid argument with a false premise, and the two criticisms are different. Calling it invalid loses the distinction that lets you locate the actual fault.
Soundness is the pair
An argument is sound when it is valid and all its premises are true. Only then is the conclusion safe to act on. That gives four combinations worth being able to name:
| Validity | Premises | What you get |
|---|---|---|
| Valid | All true | Sound. Conclusion guaranteed true. |
| Valid | One or more false | Unsound. Conclusion may be anything. |
| Invalid | All true | Unsound. Conclusion may be anything. |
| Invalid | One or more false | Unsound, and doubly so. |
The table's practical use is diagnostic. When you disagree with a conclusion, one of exactly two things has gone wrong: a premise is false, or the step does not follow. Deciding which one saves an enormous amount of unproductive argument.
The counterexample test
Here is an invalid argument with a true conclusion:
Every multiple of 4 is even. 12 is even. Therefore 12 is a multiple of 4.
The conclusion is true, since 12 is 4 times 3. The argument is still worthless, and you can show it in one move. Keep the form and change one number:
Every multiple of 4 is even. 6 is even. Therefore 6 is a multiple of 4.
Both premises are true and the conclusion is false. Since validity is a property of form, and this form permits true premises with a false conclusion, the form is invalid — which means the original argument, whatever the truth of its conclusion, never established anything.
That is the counterexample method, and it is the most useful practical tool in this article. It requires no notation, no truth tables and no vocabulary. Take the shape of the argument, substitute content you can evaluate at a glance, and see whether you can break it. The free and openly licensed textbook forall x: Calgary, part of the Open Logic Project, develops the same notions — consequence, validity, and the formal machinery for testing them — for readers who want the systematic treatment rather than the shortcut.
Deduction in ordinary life
The everyday cases are mostly eliminations and rule applications.
Elimination. A parcel is in the hallway, the kitchen or the study. It is not in the hallway. It is not in the kitchen. Therefore it is in the study. The certainty here is real, and it is entirely inherited from the first premise: if the parcel could also be in the garage, the argument collapses. Deductions in real life usually fail at exactly this point — the list of possibilities was not complete.
Rule application. A warranty says it covers manufacturing defects only. A crack caused by dropping the item is not a manufacturing defect. Therefore the warranty does not cover it. Contracts, eligibility rules and game rules are written in this style precisely so that their consequences are decidable rather than negotiable.
This is a different activity from weighing evidence about what probably happened, which is the business of inductive reasoning, and different again from assessing where a claim came from and how far it can be trusted — the subject of our guide to primary and secondary sources. A valid deduction from a badly sourced premise is still a badly founded belief.
Why believable conclusions feel valid
There is a well-documented interference between what an argument concludes and how people judge its structure. Arguments whose conclusions match what someone already takes to be true get endorsed as valid more often than structurally identical arguments with unbelievable conclusions.
A hierarchical Bayesian meta-analysis by Dries Trippas and colleagues, published in Psychonomic Bulletin and Review in 2018, pooled 22 studies covering 993 participants and reached a more specific conclusion than the older literature: believability did not appear to change people's underlying ability to discriminate valid arguments from invalid ones, but did shift their general willingness to endorse arguments. On that reading, the problem is a bias in where people set their threshold rather than a failure of the reasoning itself.
The countermeasure follows from the counterexample test. Judge the form with content you have no stake in. If an argument about a contested subject worries you, rewrite it about squares, birds or multiples of four, and see whether it still works. If it does not, the disagreement was never about the topic.
Practising this honestly
There is no arithmetic in this article, and the app's arithmetic drills are not what helps with it; its logic games, which include deduction and conditional reasoning, are the closer match. Our app also offers a set of untimed Accuracy Trainer rounds with feedback and explanations, which are the opposite of the timed sprints in one specific way: they do not reward answering before you have finished thinking. If your errors come from speed rather than from method, that pressure difference is the relevant one. The app itself is described on the IQ Test & Brain Training Games app page, and our sceptical review of what practice apps can and cannot do is in do brain training apps work.
For the quantifier machinery behind "all", "some" and "none", continue to syllogisms. For rule-discovery problems, where the premises are a handful of terms rather than sentences, start with number sequence puzzles. The rest of our writing on reasoning and practice is in the education and brain training section.
Good to know
Frequently asked questions
What is deductive reasoning in simple terms?
Is it possible for reasoning to be perfect and the answer still be wrong?
How do you prove an argument is invalid?
Where does deduction actually come up?
Sources
- Logical Consequence (opens in a new tab)Stanford Encyclopedia of Philosophy — accessed
- forall x: Calgary. A Free and Open Introduction to Formal Logic (opens in a new tab)Open Logic Project, University of Calgary — accessed
- Characterizing belief bias in syllogistic reasoning: A hierarchical Bayesian meta-analysis of ROC data (opens in a new tab)Psychonomic Bulletin and Review, via PubMed Central — accessed
About the author
Reign Creative LLC — The independent app studio that builds the apps covered on this site and publishes every article here. On Google Play it publishes as Reign Collective Apps.
We use AI tools to help research, draft and edit our articles. Each article lists the sources its factual claims rely on, and details about our apps come from their live Google Play listings. No article on this site has been reviewed by an independent expert unless the reviewer is named on it.
Spotted an error? Tell us — how corrections are handled is set out in our editorial policy.