¹ We are using the Everyman, edition of Leviathan (London: Dent, 1947), p. 64.Google Scholar

² Leviathan, p. 64.

³ Hobbes's condition of Warre is not necessarily a condition of continual fighting, ‘For WARRE, consisteth not in Battell onely, or the act of fighting; but in a tract of time, wherein the Will to contend by Battell is sufficiently known… So the nature of War, consisteth not in actual fighting; but in the known disposition thereto, during all the time there is no assurance to the contrary. All other time is PEACE.’ (Leviathan, p. 64.)

⁴ Notice that Hobbes does cater to this problem to some extent. Men will never be satisfied with a particular level of satisfaction because it can never be secure; a search for power will go on ‘that ceaseth onely in Death’. Continual attempts to secure power over others are a rational response to the fact that, in a state of nature, no man can be quite confident of his own position. ‘And the cause of this [restless desire for power], is not alwayes that a man hopes for a more intensive delight, than he has already attained to; or that he cannot be content with a moderate power: but because he cannot assure the power and means to live well, which he hath at present, without the acquisition of more’ (pp. 49–50). In these terms, no level of abundance will keep men from the war of all against all. Nevertheless, it is possible now to conceive a level of abundance, and a distribution of that abundance within the population, that would satiate all and provide the necessary confidence in the satiation of others for conflict to cease.

⁵ Leviathan, pp. 64–5.

⁶ Leviathan, p. 89.

⁷ Alker, Hayward, Mathematics in Politics (New York: Macmillan, 1965), pp. 112–29.Google Scholar

⁸ Leviathan, pp. 89–90.

⁹ In Leviathan, Hobbes makes a clear statement of the power involved in the rule-making function: ‘the whole power of prescribing the Rules, whereby every man may know, what Goods he may enjoy, and what Actions he may doe, without being molested by any of his fellow Subjects’ (p. 93). The operational characteristics of Leviathan are found in Chap, XVIII entitled ‘Of the RIGHTS of Soveraignes by Institution’.

¹⁰ The following is how Hobbes expresses the rule-adjudication function: ‘the Rights of Judicature; that is to say, of hearing and deciding all Controversies, which may arise concerning Law, either Civil or Naturall, or concerning Fact’ (p. 93).

¹¹ The following is how Hobbes expresses the rule-application function: ‘the Power of Rewarding with riches, or honour; and of Punishing with corporall, or pecuniary punishment, or with ignominy every Subject according to the Law he hath formerly made; or if there be no law made, according as he shall judge most to conduce to the encouraging of men to serve the Commonwealth, or deterring of them from doing dis-service to the same’ (p. 94).

¹² There is a certain amount of redundancy in Hobbes's list, and we have avoided duplicating it here.

¹³ The phrase attributed to John Tukey puts the argument well: ‘Anything worth doing is worth doing superficially’.

¹⁴ Banks, Arthur S. and Textor, Robert B., A Cross-Polity Survey (Cambridge, Mass.: MIT Press, 1963).Google Scholar

¹⁵ Banks, and Textor, , A Cross-Polity Survey, p. 83.Google Scholar

¹⁶ Banks and Textor distinguish three types of constitutional status: (A) constitutional (government conducted with reference to recognized constitutional norms); (B) authoritarian (no effective constitutional limitation, or fairly regular recourse to extra-constitutional powers, arbitrary exercise of power confined largely to the political sector); (C) totalitarian (no effective constitutional limitation, broad exercise of power by the regime in both political and social spheres); (p. 83). The Cross-Polity Survey, as well as the other sources we used, has some informational gaps where for some reason data were not available. In these cases, we have opted for less rigorous data rather than none; we asked colleagues who had expert knowledge of the country concerned to make the best judgment possible in terms of the criteria used by Banks and Textor.

¹⁷ The quote is from Almond, Gabriel A. and Coleman, James S., The Politics of Developing Areas (Princeton, N. J.: Princeton University Press, 1960), p. 34Google Scholar, and is cited by Banks and Textor in their discussion of ‘Interest articulation by Associational Groups’, which is on p. 89 of A Cross-Polity Survey. Banks and Textor distinguish four degrees of associational power: (A) Significant, (B) Moderate, (G) Limited, (H) Negligible.

¹⁸ Banks, and Textor, , A Cross-Polity Survey, p. 90.Google Scholar Cited from Almond, and Coleman, , The Politics of Developing Areas, p. 33.Google Scholar

¹⁹ Banks and Textor distinguish four degrees of institutional power: (A) Very significant, (B) Significant, (G) Moderate, (H) Limited.

²⁰ Banks and Textor, A Cross-Polity Survey, p. 92. Cited from Almond, and Coleman, , The Politics of Developing Areas, p. 33.Google Scholar

²¹ Banks and Textor, A Cross-Polity Survey, p. 92. The authors distinguish four degrees of non-associational group power: (A) Significant, (B) Moderate, (G) Limited, (H) Negligible.

²² Banks and Textor, A Cross-Polity Survey, pp. 92–3. They distinguish four degrees of anomic group articulation: (A) Frequent, (B) Occasional, (G) Infrequent, (H) Very infrequent.

²³ Banks and Textor distinguish four degrees of political party articulation: (A) Significant, (B) Moderate, (C) Limited, (D) Negligible.

²⁴ Hobbes was not impressed, as were later authors like Locke and Montesquieu, with the virtues of separating these powers. Unlike these two authors, Hobbes was concerned to maximize the power of government; they were concerned to limit the power of government, and, in Locke's case, to justify a revolution against it.

²⁵ Banks and Textor, A Cross-Polity Survey, p. 106. They distinguish three degrees of horizontal power distribution: (A) Significant (effective allocation of power to functionally autonomous legislative, executive and judicial organs), (B) Limited (one branch of government without genuine functional autonomy, or two branches with limited functional autonomy), (C) Negligible (complete dominance of government by one branch or by extra-governmental agency).

²⁶ The correlation model used is the conventional Pearson product-moment correlation. Although its implementation for data which are clearly ordinal (such as the present data) involves relaxing assumptions, we believe that more difficulties are involved when mixing non- and quasi-metric correlation models, particularly when multivariate analysis is needed. See: Brent M. Rutherford, ‘Non-Metric Correlation Methods; A Sensitivity Analysis Using Monte Carlo Simulation’, paper delivered at the 1971 meeting of the Pacific Sociological Association, Honolulu, Hawaii.

²⁷ A standard principle components factoring routine was utilized, specifically BMD 72. The values of successive eigenvalues were 3.41, 1.13, 0.82, 0.56, 0.46, 0.32 and 0.30. Such a pattern is strong evidence for single factorness (see Linn, R. L, ‘Monte Carlo Approach to the Number of Factors Problem’, Psychometrika, XXXIII (1968), 37–71).CrossRefGoogle Scholar Moreover, all variables load uniformly on the first factor, from the first variable to the seventh: - 0.66, - 0.72, 0.82, 0.72, - 0.72, - 0.55, and 0.65. These loadings suggest that each indicator is empirically reflective of a single underlying dimension which we judge to be the concept of Leviathan.

²⁸ Gurr, Ted Robert, ‘ A Comparative Study of Civil Strife’, Chap. 17 of Violence in America, The Official Report to the National Commission on the Causes and Prevention of Violence (New York: Bantam Books, 1969).Google Scholar

²⁹ Gurr, ‘ A Comparative Study of Civil Strife’, Appendix 1, p. 597. This appendix contains Gurr's methodology for arriving at his national scores, and we will not repeat it here.

³⁰ Rummel, Rudolph J., ‘Dimensions of Conflict Behavior Within and Between Nations’, in Bertalanffy, Ludwig Van and Rapoport, Anatol, eds., General Systems; Yearbook of the Society for General Systems Research, VIII (1963), 1–50.Google Scholar

³¹ Ivo K. Feierabend, Rosalind L. Feierabend, and Betty A. Nesvold, ‘Social Change and Political Violence: Cross-National Patterns’, Chap. 18 in Violence in America.

³² Campbell, D. T. and Fiske, D., ‘Covergent and Discriminant Validation by the Multi-trait Multi-method Matrix’, Psychological Bulletin, LVI (1959), 81–105.CrossRefGoogle Scholar

³³ Deutsch, Karl W., Nationalism and Social Communication (Cambridge, Mass.: MIT Press, 1953), p. 87.Google Scholar

³⁴ Deutsch, , Nationalism and Social Communication, p. 96.Google Scholar

³⁵ Figures on mail-flow per capita, and the remaining indicators of social output variables, are derived from Russett, Bruce M., Alker, Hayward R. Jr, Deutsch, Karl W., and Lasswell, Harold D., World Handbook of Political and Social Indicators (New Haven: Yale University Press, 1964).Google Scholar

³⁶ The successive eigenvalues were: 3.71, 1.55, 0.75, 0.67, 0.51, 0.36, 0.29 and 0.16. The decision of two factors is, again, based on the experiments by Linn, fn. 32.

³⁷ Specifically, an orthogonal verimax rotation. Although non-orthogonal rotation options are feasible, we are interested, for the moment, in maximal parsimony and separation.

³⁸ See Kuder, G. F. and Richardson, M. W., ‘The Theory of Estimation of Test Reliability’, Psychometrika, II (1937), 151–60.CrossRefGoogle Scholar

³⁹ Each indicator was standardized, i.e., transformed to a zero mean, unit deviate distribution. The commodiousness score is the sum of variables (1) through (5); the sum was then divided by five. Likewise, the standardized violence measures were summed and divided by three. If orthogonal factor scores had been utilized, the mathematical model would have imposed a zero correlation between commodity and violence, even though the factor matrices indicate the overlap exists. Thus, the mathematical model would have imposed drastic method effects in evaluating the causal linkages. Non-orthogonal factor scores might have been utilized, but the mathematical issues are sufficiently complex and the gain over simple standardization and summation so slight that we judged our scoring method adequate.

⁴⁰ Because the cases in analysis are the population of world polities and not a sample, testing for statistical significance would not fit the traditional model of statistical inference. However, as a point of reference, a correlation of ·32 is significant at the ·05 level with 111 degrees of freedom when testing a directional (one-tailed) hypothesis.

⁴¹ As the relationship between Leviathan and felicity is non-linear, the reported correlation value would be increased by appropriate variable transformations. Argument for transformation is both current and well grounded in the mathematical characteristics of the covariance model; however, if we had transformed the variables in analysis, the special clustering of the socialist polities would have been obscured, and we believe that transformation would have little to contribute to our theoretical understanding of the Hobbes hypothesis.