User Rating 0.0 β˜…β˜…β˜…β˜…β˜…
Total Usage 0 times

Write one premise per line. Use logical connectors like because, therefore, if…then, since, hence.

Enter an argument above and click Validate Argument to begin analysis.

Is this tool helpful?

Your feedback helps us improve.

β˜… β˜… β˜… β˜… β˜…

About

Logical arguments fail silently. A structurally invalid argument can sound persuasive. An argument riddled with fallacies can pass unnoticed in debate, academic writing, or legal reasoning. This tool parses natural-language arguments into their component parts - P1, P2, …, Pn (premises) and C (conclusion) - then evaluates structural validity against standard inference rules such as modus ponens and disjunctive syllogism. It cross-references over 25 catalogued formal and informal fallacies. The strength score (0 - 100) reflects structural integrity, fallacy presence, and premise verifiability.

Limitations: The tool approximates natural-language parsing. It cannot verify empirical truth of premises. Ambiguous phrasing or implicit premises may reduce detection accuracy. For best results, state each premise on its own line and use explicit logical connectors (because, therefore, if…then). This is a structural analysis tool, not a truth oracle.

argument validator logical fallacy detector validity checker critical thinking propositional logic argument analysis reasoning checker

Formulas

The argument strength score S is a weighted composite of three dimensions:

S = 0.30 β‹… Sstruct + 0.40 β‹… Sfallacy + 0.30 β‹… Spremise

Where Sstruct = structural quality score (0 - 100). Awards points for: identifiable conclusion (30 pts), at least 2 premises (30 pts), logical connector usage (20 pts), and valid inference pattern match (20 pts). Sfallacy = fallacy-free score. Starts at 100 and subtracts 20 per detected fallacy, floored at 0. Spremise = premise support score. Evaluates whether premises contain verifiable claims vs. subjective assertions. Each verifiable premise adds 100n where n = total premise count.

Validity check applies pattern matching against standard inference rules:

Modus Ponens: P β†’ Q, P ∴ Q
Modus Tollens: P β†’ Q, Β¬Q ∴ Β¬P
Hypothetical Syllogism: P β†’ Q, Q β†’ R ∴ P β†’ R
Disjunctive Syllogism: P ∨ Q, ¬P ∴ Q

Reference Data

FallacyTypePatternExample
Ad HominemInformalAttacks the person, not the argument"You're wrong because you're biased."
Straw ManInformalMisrepresents the opposing position"So you're saying we should have no rules at all?"
Appeal to AuthorityInformalCites authority outside relevant expertise"A celebrity says this diet works."
False DilemmaInformalPresents only two options when more exist"Either you support X or you support chaos."
Slippery SlopeInformalAssumes chain of events without justification"If we allow A, then B, C, and D will follow."
Circular ReasoningInformalConclusion restates a premise"It's true because it's a fact."
Red HerringInformalIntroduces irrelevant topic to divert"Why worry about X when Y is happening?"
Appeal to EmotionInformalUses emotional manipulation over evidence"Think of the children!"
BandwagonInformalAppeals to popularity as proof"Everyone believes this, so it must be true."
Hasty GeneralizationInformalConcludes from insufficient sample"I met two rude locals, so everyone there is rude."
Tu QuoqueInformalDeflects by accusing hypocrisy"You do the same thing!"
Appeal to NatureInformalAssumes natural equals good"It's natural, therefore it's safe."
Appeal to IgnoranceInformalClaims truth from lack of disproof"No one has proven it wrong, so it's true."
EquivocationInformalShifts word meaning mid-argument"The sign said "fine for parking here." So it's fine to park."
Loaded QuestionInformalPresupposes unproven assumption"When did you stop cheating?"
No True ScotsmanInformalRedefines group to exclude counterexamples"No real expert would disagree."
Post HocInformalAssumes causation from sequence"After the policy, crime dropped - so the policy worked."
Affirming the ConsequentFormalP β†’ Q, Q ∴ P"If it rains the ground is wet. The ground is wet, so it rained."
Denying the AntecedentFormalP β†’ Q, Β¬P ∴ Β¬Q"If it rains the ground is wet. It didn't rain, so the ground isn't wet."
Undistributed MiddleFormalMiddle term not distributed in syllogism"All dogs are animals. All cats are animals. So all dogs are cats."
Illicit MajorFormalMajor term distributed in conclusion but not premise"All roses are flowers. No daisies are roses. So no daisies are flowers."
False CauseInformalIncorrectly identifies causal relationship"Ice cream sales and drowning correlate, so ice cream causes drowning."
CompositionInformalAssumes whole has properties of parts"Each brick is light, so the wall is light."
DivisionInformalAssumes parts have properties of whole"The team is great, so every player is great."
Genetic FallacyInformalJudges by origin instead of content"That idea came from a flawed study, so it's wrong."

Frequently Asked Questions

The parser scans for conclusion indicators (therefore, thus, hence, so, consequently, it follows that, which means) and premise indicators (because, since, given that, as, for, assuming that). Sentences containing conclusion indicators are tagged as conclusions. Remaining declarative sentences are treated as premises. If no indicators are found, the last sentence defaults to the conclusion role.
Only partially. The fallacy engine relies on explicit textual patterns - keywords, phrase structures, and sentence relationships. If a premise is implied but never stated, the tool cannot analyze it. For best results, make all premises explicit. Write one premise per line.
Validity is structural: if the premises were true, would the conclusion necessarily follow? An argument can be valid with false premises. Soundness requires both valid structure and true premises. This tool evaluates validity (structural form) but cannot verify empirical truth of premises. The premise support score approximates this by flagging subjective or unfalsifiable claims.
The score penalizes heavily for detected fallacies (40% weight). A structurally sound argument containing even one fallacy loses 20 points from the fallacy subscore. Two fallacies reduce that component to 60 out of 100. Additionally, premises containing opinion words (should, believe, feel, obvious) lower the premise support score.
Moderately. The parser uses keyword indicators regardless of order, so "Therefore X because Y" works correctly. However, arguments structured as one premise per line with the conclusion last yield the most reliable parsing. Avoid compound sentences containing both premises and conclusions separated only by commas.
No. This tool operates on natural-language propositional arguments. Formal proofs require symbolic logic systems with axiom schemas and derivation rules beyond pattern matching. Use this tool for everyday reasoning, essays, debates, and informal logic assessment.