Journal of the History of Philosophy

Hume's Criticism and Defense of Analogical Argument CATHERINE S. FRAZF.R LIKE MOST HISTORIANS,David Hume relied extensively upon analogical arguments in his History of England. For example, in studying the origin of various tribes inhabiting Scotland and England, Hume writes that: It appears more than probable, from the similitude of language and manners, that Britain either was originally peopled, or was subdued, by the migration of inhabitants from Gaul, and Ireland from Britain.... It appears also probable, that the migrations of that colony of Gauls or Celts . . . was originally made from the northwest parts of Britain; and this conjecture (if it do not merit a higher name), is founded both on the Irish language.., and from the language anciently spoken in South Britain.... These events, as they passed long before the age of history and records, must be known by (analogical) reasoning alone, which in this case seems to be pretty satisfactory. Caesar and Tacitus, not to mention a multitude of other Greek and Roman authors, were guided by like inferences,x Hume was also relying upon analogy when he observed that "the similarity of these institutions [of late Saxon government] to the customs of the ancient Germans, to the practices of the older northern conquerors, and to the Saxon laws during the Heptarchy, prevents us from regarding Alfred as the sole author of this plan of government.... ,, 2 Again, this is Hume's method in criticizing King Henry's treatment of Beckett, as he argues from the well-established fact of the king's precipitous recall of Beckett's financial accounts to a presumption of similar misconduct about the document purportedly containing Beckett's admission of perjury: "If the king was so palpably oppressive in one article, he may be presumed to be equally so in the rest." 3 Such examples are numerous. In some cases, the professional historian today could challenge the accuracy of the specific details which Hume cites. This task, however, is less interesting to the philosopher, whose concern is primarily with the structure and validity of the method of analogical reasoning as such. Hume's use a Dsvid Hume, The History o/ England (Edinburgh- Printed by Oliver & Co. for T. Brown, Bookseller, Northbridge, and T. Oliver, Netherbow, 1830), tight volumes, Vol. I, p. 24, n. " Ibid., I, 81. " Ibid., I, 291, n. See also II, 84, 93; VI, 180; VIII, 63. [173] 174 HISTORY OF PHILOSOPHY of analogy in history writing is of special interest in view of his searching criticisms of analogical argument in his philosophical works, especially in the Dialogues Concerning Natural Religion. Was Hume the historian simply turning his back on the skeptical doubts he had raised as a philosopher? Or does his philosophical position contain a ddense of analogical reasoning that could justify his reliance upon such methods of argument eisevehere? The problem of Hume's consistency on this question of analogy is reminiscent of the problem of interpreting his attack on causality. Many have argued that Hume defiled all possibility of causal inference. But a more caxeful reading suggests that on both these points of causal and analogical inference, Hume provided his own .resolution of the skeptical doubts he had raised. An argument is said to be analogical when the line of reasoning runs from known characteristics of an object or event to the conclusion that a second object, in some ways similar to the first, exhibits other like characteristics. We base our assurance about such conclusions upon the fact of the observed similaxity. Even though we do not directly know the farther characteristics of the second object, we infer that they do occur by virtue of the observed similarity between the objects in question.4 The line of argument is necessarily indirect. Had Hume been able to find direct evidence concerning these early migrations, there would have been no need for such indirect reconstruction. Whenever such direct data is lacking, the analogical argumem is the historian's only way to bridge the gap between what he can observe and what cannot .be observed directly. Hume believes that such a procedure can often be successful. One might extend the argument further, by pointing out that historical inferences always require...