Second art of the trivium. Part of The Cottonwood Collection — a public reference library on harm, care, and stewardship.
Aristotle’s Organon (4th c. BCE) — comprising the Categories, On Interpretation, Prior Analytics, Posterior Analytics, Topics, and Sophistical Refutations — establishes logic (analytika) as the instrument for scientific demonstration. A valid argument is a syllogism: a deduction in which, certain things being stated, something other than what is stated follows of necessity from their being so. The classic form is: All B is A; All C is B; therefore, All C is A. Proof (apodeixis) is a syllogism producing scientific knowledge from premises that are true, primary, immediate, better known than, prior to, and causes of the conclusion (Posterior Analytics). Validity is a matter of form, independent of content.
The Megarian school (Euclides of Megara, 4th c. BCE) and later the Stoics (especially Chrysippus, 3rd c. BCE) developed propositional logic. They analyzed arguments using connectives like “if…then” (conditional), “or” (disjunction), and “and” (conjunction). A valid argument is one where the contradictory of the conclusion is incompatible with the conjunction of the premises. The Stoics identified five basic indemonstrable (anapodeiktoi) argument forms, akin to inference rules — for example: “If it is day, then the sun is shining. It is day. Therefore, the sun is shining.”
Plato’s Gorgias and Phaedrus (4th c. BCE) present a foundational tension: rhetoric as mere persuasion versus philosophy as truth-seeking. Aristotle’s Rhetoric partially reconciles this, treating rhetoric as the counterpart (antistrophos) of dialectic. Logic (logos) provided the structure for dialectical training, while rhetoric taught its persuasive application in public life. The two were intertwined in Hellenistic and Roman education, with logic training the mind for precise argumentation.
Gautama’s Nyaya Sutras (c. 2nd c. BCE – 2nd c. CE) systematize logic (nyaya, “rule” or “method”) as one of sixteen categories within a broader epistemology. Valid knowledge (prama) arises from reliable means of knowing (pramana): perception (pratyaksha), inference (anumana), comparison (upamana), and verbal testimony (shabda). A valid inference requires a five-member syllogism: proposition (pratijna), reason (hetu), example (udaharana), application (upanaya), and conclusion (nigamana). The classic illustration: “There is fire on the hill (pratijna), because there is smoke (hetu), as in the case of a kitchen (udaharana); the hill has smoke like the kitchen (upanaya); therefore, the hill has fire (nigamana).” The core logical form is based on pervasion (vyapti), an invariable concomitance between the reason (hetu) and the property to be proven (sadhya) observed in examples.
Dignaga’s Pramanasamuccaya (5th c. CE) and Dharmakirti’s Pramanavarttika (7th c. CE) refine logic within a Buddhist metaphysical framework. They reduce valid pramanas to perception and inference. Dharmakirti defines a valid reason (hetu) by three conditions: it must be present in the subject (paksa), present in similar instances (sapaksa), and absent from dissimilar instances (vipaksa). Proof is established through essential nature (svabhava) or causal relation (tadutpatti). This system engaged in intense debate with Nyaya realists over the nature of universals and the validity of scripture.
Gangesha Upadhyaya’s Tattvacintamani (14th c. CE) initiates the “New Logic” (Navya-Nyaya), shifting focus from the psychology of inference to a formal analysis of epistemic relations. It develops a technical language to precisely define terms like avacchedaka (limitor) and nirupya (qualificand). Validity is analyzed through abstract relational structures, making it a sophisticated tool for philosophical analysis across Indian schools.
The Mohist Canons (Mo Jing, c. 4th–3rd c. BCE) of the Later Mohist school develop a systematic logic centered on the relationship between names (ming) and objects (shi). They classify names into universal (da ming), classifying (lei ming), and private (si ming). A valid argument involves “picking out” (qu) an object with a name and “adducing” (shuo) a reason. They analyze argument forms, including deduction (tui), analogy (mou), and parallel (pi). Proof is established by following “patterns” (fa) or standards. The Xiaoqu (“Small Pick”) section provides further analysis of analogical reasoning.
The School of Names (Ming Jia), exemplified by Gongsun Long (c. 320–250 BCE) in his essay “On the White Horse” (Bai Ma Lun), explores paradoxes arising from the rigidity of names. His argument “a white horse is not a horse” distinguishes between the compound name “white horse” (which includes the attribute ‘white’) and the simple name “horse.” This highlights tensions between identity and attribute, challenging simple correspondence between ming and shi.
The Later Mohists advanced analogical reasoning, defining it as “comparing items and presenting them together in parallel” (Mo Jing A70). They also developed a nascent concept of logical necessity, distinguishing between “so-of-itself” (gu) and “cause” (gu written with a different character). Their logic remained embedded in ethical and political discourse, aimed at “benefiting all under heaven.”
Al-Farabi (c. 872–950) in his Commentary on Aristotle’s De Interpretatione and Book of Letters positioned logic (mantiq) as the universal “instrument” (ala) of thought, preceding all sciences. He integrated Aristotelian and Stoic elements. Ibn Sina (Avicenna, 980–1037), in al-Shifa (The Healing) and al-Isharat (Pointers), developed an original modal logic, analyzing temporalized necessity and possibility, and introduced the concept of “conditional” (sharti) syllogisms.
Al-Ghazali (1058–1111), despite his critique of Hellenistic philosophy in The Incoherence of the Philosophers, later endorsed logic as necessary for jurisprudence in The Standard of Knowledge. Ibn Rushd (Averroes, 1126–1198) defended demonstrative reasoning (burhan) in The Decisive Treatise, arguing for its harmony with Islamic revelation. Logic was integrated into the methodology of Islamic jurisprudence (usul al-fiqh), particularly in analogical reasoning (qiyas), where a ruling is extended from an established case to a new one based on a common effective cause (‘illa).
Boethius (c. 477–524) transmitted and translated Aristotle’s logical works (logica vetus) in Latin, alongside Porphyry’s Isagoge — an introduction to Aristotle’s Categories that shaped medieval debates about universals. Peter Abelard (1079–1142), in Dialectica and Logica ‘Ingredientibus’, advanced propositional logic and the theory of entailment (consequentia), analyzing the truth conditions of conditionals. William of Ockham (c. 1287–1347), in Summa Logicae, developed a nominalist logic, emphasizing supposition theory (suppositio) — how terms stand for things in propositions. His principle of parsimony (“entities must not be multiplied beyond necessity”) became a foundational heuristic for scientific reasoning.
The scholastic method of quaestio and disputatio was applied logic in action. A valid argument was one that could withstand objection and resolution in a formal disputation. Proof, particularly in theology, often relied on syllogistic reasoning from scriptural or patristic authorities, but the method itself trained a rigorous, analytic approach to all subjects.
Gottlob Frege’s Begriffsschrift (1879) created modern formal logic with a complete system of quantificational logic, divorcing logical form from natural language. Bertrand Russell, with Alfred North Whitehead in Principia Mathematica (1910–1913), aimed to reduce mathematics to logic. Kurt Godel’s incompleteness theorems (1931) demonstrated inherent limitations to this formalist program, proving that any consistent formal system capable of arithmetic contains true statements that cannot be proven within the system.
In Tractatus Logico-Philosophicus (1921), Ludwig Wittgenstein presented logic as the transcendental scaffolding of the world and language, arguing that what can be said at all can be said clearly. In Philosophical Investigations (1953), he rejected this formalist picture, arguing that logic is grounded in ordinary language games and forms of life, shifting focus from ideal structures to actual use.
Paraconsistent logics (e.g., developed by Newton da Costa, 1960s) reject the principle of explosion (ex contradictione quodlibet), allowing for coherent reasoning with contradictory premises. This has direct relevance to traditions like Buddhism (with its use of four-cornered negation, catuskoti) and Jain logic (syadvada, “maybe” theory), which systematically tolerate or incorporate apparent contradictions without rejecting the entire logical system.
Agreement on Core Structure: Greek syllogistic, Nyaya inference based on vyapti, and Islamic qiyas all share a triadic structure: two premises establishing a universal relationship lead to a conclusion about a particular case. Validity is often seen as a matter of form or invariable relation.
Divergence on the Grounds of Necessity: What secures the universal premise differs fundamentally. For Aristotle, it is essential nature grasped by intellect; for Nyaya, it is vyapti observed through positive and negative examples; for Dharmakirti, it is a real connection in reality (svabhava or causality); for the Later Mohists, it is the matching “pattern” (fa).
The Role of Contradiction: The Aristotelian (and mainstream Western) principle of non-contradiction is absolute. In Jain syadvada, multiple, seemingly contradictory predications about a complex object from different perspectives are all valid. Paraconsistent logic provides a formal framework for such non-Aristotelian approaches.
The Purpose of Logic: In Greece and the medieval West, logic was primarily an instrument for attaining certain truth (episteme, scientia). In India, it was a pramana for reliable knowledge within soteriological debate. In China, Mohist logic was a tool for social and ethical order (bian, “disputation,” to distinguish right from wrong). These differing ends shaped the development and emphasis of each tradition’s logical systems.
This page is incomplete. The following are known gaps we have not yet addressed:
If you can help fill these gaps, the source repository is open.
This page was generated by the Cottonwood Research System — multiple AI providers contributing research in parallel, synthesized into a single reference document. Raw provider responses are preserved in the source repository for full traceability.