The connection between knowledge and action can be examined empirically in two ways: the first by studying the relationship that knowledge has to action and the second by examining the connection between action and knowledge. Without trying to include the whole span of the problems involved in this double link, this article attempts to sketch some aspects of the impact which action claims to have upon knowledge by examining certain practices which emerge in an explicit way from such a connection. Set in the context of Canada and Quebec, the examination focusses chiefly on some governmental reforms which try to establish explicit functions of control, coordination, and planning. The theme of these reforms exhibits the desire on the part of these governments to stress the deliberate, accountable, and rational nature of the decisions. It seems to be a question of a new compulsion to want to break with the opportunism of the past and to reestablish the links between the “scholar” and the “politician.” In this situation how well do the reforms come out? Is there a real connection with knowledge or is it alluded to only to justify the political pronouncement? These are the questions on which the author tries to shed some light.

The systematic investigation of party-projected images is a relatively recent development in political science. Past studies into this subject have largely dealt with the perceptions of political parties held by various groups, and the perceived mobility of the parties over time with regard to selected dimensions. For the determination of party images held by individuals, most previous works in the field have made use of the semantic differential technique pioneered by Osgood, Suci, and Tannenbaum. This approach consists of respondents being asked to rank political parties along a number of seven-point scales, each demarcated by a pair of antonyms such as “good-bad” or “slow-fast.” Tests confirming the reliability, validity and comparability of the measure have also been conducted by Osgood et al., as found in The Measurement of Meaning.

Le but de cette étude est d'analyser et de décrire statistiquement la violence collective en Ontario et au Québec entre 1963 et 1973. Il y a un cas de violence collective dans la perspective de cette recherche, chaque fois qu'au moins cinquante individus (« insurgés » aussi bien que membres des forces de l'ordre) infligent des dégâts à des personnes ou à la propriété. Nous avons exclu les dommages rattachés aux jeux comme le rugby, le hockey et les courses de voiture (où des règles déterminent les modalités et l'intensité de la violence) de même que les dégâts psychiques. Notre étude se limite donc à la violence collective dite « ouverte », ce qui par conséquent élimine de notre recherche les cas de terrorisme.