Chapter 93
Reasoning and argument; inference, demonstration and the use of reason.
620 passages · 178 principal · Covers the 49 volumes of primary texts
Reasoning, as the idea is understood here, covers the whole business of passing from what is known to what is not: inference and demonstration, argument in the schools and in the street, and the faculty that does this work. It is a great idea for these books because nearly every author in them either reasons or reflects on reasoning. Socrates defends himself by cross-examination. Harvey overturns Galen by it, and Hume asks whether our commonest inferences rest on it at all. Pascal at once relies on reason and humbles it. The same questions keep returning. What reason is, how it should proceed, where it goes wrong, and how far it can be trusted against imagination, passion and faith are never settled in these pages.
The most austere definition belongs to Hobbes. He calls reasoning reckoning, the adding and subtracting of the consequences of agreed general names, whether in arithmetic, geometry, logic, politics or law1. He notes that the Greek *logos* signifies both speech and reason2. For him reason is tied to words. It must proceed step by step from first definitions, and conclusions taken on the trust of authors yield belief, not knowledge3. Man alone can reduce consequences to general rules, and for that very reason he alone is subject to absurdity, chiefly from failing to begin with definitions4. Metaphors and the shifting names of virtues and vices are therefore unfit for ratiocination5,6. Reason is “the pace,” science “the way,” and the benefit of mankind the end7.
Marcus Aurelius sees in reason something larger than a calculus. Reason common to all men implies a common law and a common citizenship of the world8. It is self-sufficient, moving from its own first principle toward its end9, and whoever follows it in all things is tranquil, active and collected10. Later writers keep this sense of a shared light. Penn likens reason to the sun, common to all, and traces disagreement to men not examining by the same light11. Emerson holds that a reason for one is a reason for every other12. Descartes calls reason a universal instrument, unlike bodily organs fitted to particular acts13. Locke calls it the highest faculty, whose right improvement is the highest perfection a man can attain in this life14. Helmholtz makes it the faculty of discovering laws, so that nature appears as a kosmos answering to our logical thought15.
Whether this faculty is born or made divides the authors. Hobbes insists reason is acquired, not innate, and that men reason alike when they start from good principles16. Locke would reason with children early, in few plain words, as rational creatures17. Aesop’s crow, raising the water with pebbles, shows something like reasoning in a beast18. Descartes replies that animals show skill only in particular actions19. Hume will later turn the child into evidence against reason’s claims. Rousseau doubts that the species could have survived if preservation had waited on individual reasoning20.
The relation of reason to imagination and feeling is contested from both sides. Shelley makes reason the analytic enumeration of known relations, standing to imagination as instrument to agent21, and warns that the calculating faculty unmitigated does harm22. Hume answers that good sense is requisite even to taste, for a poem is itself a chain of reasonings23. Burke adds that the swiftness of taste proves no separate faculty, since reasoning too grows quick with practice24. Pascal goes further still and calls love a precipitate thought that remains a reason25. Sainte-Beuve, by contrast, rejects the Latin theory that subordinates imagination to reason and refuses to equate reason with a poetic genius that expresses passion26. In science the struggle is reversed. Darwin says reason must conquer imagination and common sense before the gradations of the eye make his conclusion credible27,28. Mill found that analysis dissolved prejudice yet weakened feeling29.
If reason is a faculty, it can be used well or ill, and the moderns sought rules. Descartes resolved to sweep away his opinions and admit only what survived reason’s scrutiny30. He replaced the many rules of logic with a few rigid precepts31, chief among them to proceed from the simplest objects to the complex with complete enumerations, as geometers do32. In this way each truth found becomes a rule for finding the next33. Pascal states the ideal more exactly: prove every obscure proposition from evident maxims or from propositions already demonstrated34. He also sees its limit. A perfectly complete order is impossible, since every proof presupposes earlier propositions35. The best order neither proves everything nor nothing36. Against the scholastics he holds that true rules are simple and natural, not Barbara and Baralipton37. Locke, likewise, thinks right reasoning owes little to predicaments and predicables38. He would teach it by reasoning over cases, warning that captious disputation teaches only wrangling39,40.
Behind these rules lies the older Socratic preference for argument over authority. In the *Crito*, Socrates will be guided by reason alone, honouring his principles unless better ones are found41, and he invites refutation42. Luther asks that the laity’s arguments be answered with reasons rather than force43. Hugo prefers reasons to authorities44. Yet reasoners differ. Hobbes warns that even the ablest err, and that appeals to “right reason” often merely dress up the passions45. Descartes rejected all former demonstrations because men err even in simple geometry46. Pascal treats false reasoning as a malady with remedies47. Hume catalogues the causes of difference: attention, comprehensiveness, the length of chains of inference48,49.
The forms of inference themselves range widely. Deduction is honoured in Descartes’ proof that the more perfect cannot proceed from the less50 and in Augustine’s stepwise argument that evil cannot be a substance51. Mill, however, raises its paradox: how general reasoning discovers new truth when every syllogism’s conclusion lies in its premises52. Induction and analogy appear everywhere. Socrates argues from opposites such as sleep and waking53. Simmias then turns analogy against him with the lyre and its harmony54, and Cebes shows by counter-analogy that an inference is fallacious55. Harvey argues from kidneys and skin that blood may permeate the lungs56. Lyell uses census registers and buried cities to show that abrupt fossil breaks do not prove sudden change57. Yet Lyell also warns that analogy from observed processes to unseen origins can mislead58, and Darwin concedes that analogy may be a deceitful guide59. Pascal states a principle of method here: a general theorem is shown through a particular case, since the proof is clearer than what it proves60.
Refutation by absurdity and elimination is equally old. Nebridius’s dilemma traps the Manichees61. Pascal refutes indivisibles by drawing out their absurd consequences62. Philonous rests his case on the principle that no doctrine leading to absurdity can be true63. Harvey shows by elimination that the blood swelling below a ligature can come only by the arteries64. Darwin excludes sediment and uplift to conclude that coral land has subsided65.
Probable inference from signs runs from tragedy to science. Electra reasons from hair and footprints to Orestes while admitting uncertainty66. Herodotus argues that Priam would have surrendered Helen had she been in Troy67. Cellini infers poisoning from a convergence of circumstances68. Plato’s Socrates is wary of all this: arguments from probability are impostors, unlike demonstration69. The moderns are kinder. Hume reckons more than half of human reasonings to be inferences of this kind, with certainty proportioned to experience70.
Bad reasoning has its own anatomy. Socrates exposes Meletus by showing that belief in spirits implies belief in gods71,72. Herodotus shows that the Ionian division of the earth is inconsistent73. Charges of circularity recur. Philonous accuses Hylas of begging the question74. Rousseau exposes the circle of proving doctrine by miracle and miracle by doctrine75, a circle Pascal tries to escape by distinguishing the times76. Darwin finds the same fault in using sterility to define species77.
Subtler is rationalization, reason in the service of desire. Franklin observes wryly that being a reasonable creature lets one find a reason for whatever one wants78. Hobbes states the fool’s argument that breaking covenant may be reasonable and then refutes it79,80. Milton’s Eve reasons her way past the prohibition81, and Bunyan’s Money-love builds a sophistical chain to show that gain by religion is good82. Sydney Smith classifies the political fallacies of personality, delay and vague generality83,84,85. Worst of all are the “anti-rational” fallacies that brand any plan theoretical in order to avoid discussion, for to disclaim theory is to disclaim thought86,87. Popular delusion shows reason disordered rather than absent. Quixote makes enchantment unfalsifiable88,89. Manzoni’s learned men pile up authorities90, and their elaborate reasoning from false premises ends in denying the plague91. Pascal diagnoses the habit: people reason that a thing is possible, therefore it is92.
Science and affairs show reasoning at its most productive. Harvey’s case for circulation is a sustained argument. He sets out what must be proved93, calculates that the blood expelled exceeds anything food could supply94, and concludes from argument and ocular demonstration that blood moves in a circle95. Along the way he turns Galen’s own concessions against him96. Darwin calls his book one long argument, stating objections at full force97. Pasteur refutes Fremy’s hemi-organism by showing that media without albumin still produce ferments98. Law and economy use the same tools. Marshall holds Maryland to consistency in the meaning of “necessary”99. Pliny argues from the wording of a division formula100. Smith grants a rival’s supposition and derives a contradiction from it101.
Hume takes up the deepest question, whether inference from experience is reasoning at all. Every inference to absent fact presupposes causal connexion102. Yet the step from like sensible qualities to like secret powers is supported neither by demonstrative nor by moral reasoning103. To call it experimental begs the question104. Since infants draw the inference, it cannot rest on intricate argument105. Such inferences are effects of custom, not reason106, though common life goes on undisturbed107. His treatment of testimony follows from this. Testimony is weighed like any other constant conjunction, with contrary evidence producing a mutual destruction of arguments108. Hence his maxim: no testimony establishes a miracle unless its falsehood would be more miraculous than the fact109. Pascal had held that no reasonable belief can stand against miracles110. Rousseau, on the other side, holds that the testimony of nations cannot prove an absurdity against reason’s eternal truths111. On arguing from effects to God, Hume allows the inference for human works because we know men from wider experience112. Philosophers, he says, return from the cause to new effects on imagination’s wings113. Berkeley, by contrast, thinks the inference from the sensible world to an infinite Mind needs no subtle reasoning at all114.
These doubts reopen the oldest worry. In the *Phaedo*, Socrates warns against misology, the hatred of arguments that arises, like misanthropy, from trusting arguments that later prove false115. Hume shows sceptics turning reason against itself until it is left confounded116. Kant observes that experienced users of reason come to resent it117. FitzGerald’s Khayyam divorces barren logic for the grape118.
The religious writers press reason’s weakness differently. For Augustine, abstract reasoning is too weak, so Scripture’s authority is needed119, and Thomas à Kempis would have natural investigation follow faith120. Browne finds his reason made pliable to faith by nature’s obscurities121. Pascal is the most divided. Reason is blown about by imagination122, yet its inability to prove first principles should humble it without undermining certainty123. Its last step is to recognize infinite things beyond it124. Reason can decide nothing on God’s existence125, so he calculates the stakes of a forced choice126. Against this humility Rousseau insists that reason must judge even when to submit127. Lessing makes revelation and reason teachers of one another128.
With the passions the claim of rule weakens over time. Epictetus lets reason check lust129, and Milton’s Abdiel holds that whoever wins in debate should win in arms130. Shakespeare’s Macbeth, however, finds that passion outran “the pauser, reason”131. Pope reduces reason to a guard rather than a guide, often a pleader132. Pascal says reasons follow inclination133. Schiller has reason find the law while will and feeling must execute it134.
Kant offers a final reconciliation and a final limit. Speculative and practical reason are one reason in different applications135, and reason alone can be practical and master inclination136. Yet reason ceaselessly seeks the unconditioned. It can comprehend only the incomprehensibility of practical necessity137, and to explain how pure reason can be practical lies beyond it138. So the question these books leave open is not whether men must reason; Burke notes that reasoning about practice cannot be avoided139. It is whether reason, which judges everything, can also judge itself.
Introductory essay written by Claude Opus 5.5 from the outline and the notes on every passage below; quotations are checked against this edition.
See also: Justice 1
See also: Mathematics 2b
See also: Government 1d · Progress 6a
See also: Experience 9b · Geology 6c · Science 3b · Medicine 10a
See also: Experience 9b · Medicine 10a · Animal 1d
See also: Constitution 6c · Law 9e · Government 7c
See also: Experience 2c · Cause 2a · Custom and Convention 1b · Philosophy 4c
See also: Miracles 5a · Experience 3c · Judgment 5c · Religion 3c
See also: Cause 2e · Nature 4d · God 1c
See also: Knowledge 5a · Memory and Imagination 5a · Religion 3a · Truth 2b
See also: Chance 7d · Religion 3b · God 2d
See also: Metaphysics 1b · Duty 1a · Will 1b · Liberty 1b
See also: Mathematics, Experience, Cause, Knowledge, Logic, Hypothesis