Analysis of the Canadian political system has suffered from a relative paucity of competing interpretations of the same phenomena. Too many interpretations of our polity have gone unchallenged, probably on the assumption that our scarce academic resources should not be wasted on internecine controversy while virgin fields remain untapped. Professor Lovink's article is a hopeful indication that this stage of disciplinary immaturity is ending. His sophisticated dissection of my previous article is a helpful contribution to the discussion of the effects of the electoral system on the party system. These comments, by concentrating on some of the problems raised by Lovink, are designed to contribute to a further clarification.

Initially, it can be noted, that the disagreement between us is not over the data dealing with votes and seats, but with the interpretation to be given the data. It was perhaps in the very nature of a somewhat polemical article attacking the “conventional wisdom” that I stressed the effects as I saw, or deduced, them, and in the nature of Lovink's rejoinder that the possible effects are minimized.

This difference is noteworthy in our respective treatment of the electoral system's systematic bias against Conservative Quebec voters. The data indicate that the ratio of 5.6 Liberals to 1 Conservative mp resulted from a ratio of 1.9 Liberals to 1 Conservative voter. This struck me as pregnant with consequences for the party system, some of which I tried to suggest.

Elections as the Product of Collective Decision: Federal Elections from 1957 to 1965 in Quebec

Is it fruitful to consider electoral results as the product of collective decisions, as suggested by Vincent Lemieux in his study of provincial elections in Quebec? In order to answer this question, the authors applied the same approach to new data, the results of federal elections in Quebec from 1957 to 1965. It is first noted that the operational concept of collective decision should be linked to a theoretical one. It is suggested that this theoretical concept would be more meaningful if a structural approach was adopted: it could be an “inconscient collectif” influencing the ridings in one way or another. In this sense the ridings are facing many choices: they may tend to choose a certain party, to vote for the party in government (or the contrary), or they may wish to re-elect the same party (or the contrary); finally, there might be a tendency to vote like the majority of the ridings (or, once again, the contrary). The comparison between expected and actual frequencies of the different sets of partisan choices argues in favour of the importance of party loyalty. The cleavage between the socio-economic characteristics of the traditionally Liberal, Social Credit, and Conservative ridings is also evident. The analysis of the other choices raises many problems, however. The same electoral result may refer to many different sets of choices and the relationships between these choices make it almost impossible to measure the influence of eac structural mechanism controlling for all the others. It becomes almost impossible to discover which structural law does in fact govern the ridings’ behaviour. On the whole, then, structural and causal studies are faced with the same basic methodological problem: when one wants to measure the sole impact of a particular causal factor or structural mechanism which interacts with many others, it is necessary to make some assumptions, like linearity and additivity, which exclude any complex interaction.

Anti-separatism and messianism in Quebec since 1960

Confronted by a political debate now omnipresent in Quebec society, observers have naturally focused their attention – academic or not – on the content and proponents of the “séparatiste” solution to the current Canadian dilemma. Our contention is that it takes two sides to have a debate and that some light could be thrown on the intellectual history of French Canada by reversing the perspective and focusing instead on those who oppose independence for Quebec: the anti-separatists.

Although widely diversified as to the specifics of their opposition and as to their own solution to the problem, anti-separatists nevertheless share certain ideological traits in their portrait of the French-Canadian Homus and in the characteristics and roles which they assigned to the French-Canadian collectivity. Individually, French Canadians are judged to be weak, unstable, verbose, and un-democratic; as a collectivity they are perceived as fulfilling all the requirements of a “chosen” people whose mission it is to reconquer Canada through the strength of their intellect and to show the world an example of binational co-operation.

This schizophrenic vision of French Canada is hypothesized to be a secularized version of the traditional nationalist vision first enunciated by Garneau and Parent and later developed by Groulx, Barbeau, and others. The persistence of this messianic orientation in the intellectual history of French Canada is tentatively explained through the contributions of the sociology of utopia (French Canada as an aborted utopia) and of the sociology of colonial development (French Canada as a colonial society).

This article presents a theoretical analysis of the turnover and tenure of the members of the Canadian House of Commons from 1867 through 1968. Previous studies of such turnover and tenure have been quantitative, but not theoretical, yielding precise measurements but no pattern. A statistical résumé is no substitute for a mathematical model. Both may be accurate, parsimonious, and elegant; but a mathematical theory is distinguished by its generality and its explanatory or predictive power. With the present model, for example, the members’ median and mean continuous service can be logically derived from the mathematical theory; but the converse is not true, that is, the mathematical theory cannot be deduced from the median and/or mean continuous service. Specifically, the theory implies that the median continuous service is approximately 0.693 times the mean continuous service. Despite a plethora of quantitative studies of legislative turnover and tenure, this equation (so far as we know) has not previously been discovered.

The process to be modeled can be abstractly characterized as follows: Consider the members of the House of Commons after some particular (zero) general election. These legislators are called the original members, for expository convenience. With the occurrence of deaths, resignations, political defeats, etc., only some of the original members will continue to be members of the House of Commons after the next (first) general election. These survivors are called the re-elected members.