No abstract

In recent years, we have seen a multiplication of studies estimating the location of political parties on the left-right axis using textual data. The addition of computer-assisted content analysis techniques to the researcher's toolbox has opened new fields of inquiry. So far, studies have focused on one specific language, and little attention has been paid to linguistic differences in party manifestos that are provided to the voters in more than one language. There is no theoretical reason to assume that two party platforms that have the same content but are written in different languages will not locate parties the same way on the left-right axis. However differences between languages are well documented in the field of linguistics. In the case of Canada, there are numerous differences between its two official languages, French and English, that may modify the results of textual analyses depending on which language they are performed. The relative frequency of words in a text is an example: French has gender categories (masculine-feminine) for nouns and adjectives and a more differentiated verb termination system (usually 5 compared to 2 in English). Considering the fact that quantitative content analyses are often based on word frequencies, a comparative, multi-language analysis is a first step to test the impact of linguistic specificities on party positioning. Using Wordscores and Wordfish, plus expert surveys as a benchmark, we compare the spatial location of major Canadian parties on the left-right axis using English and French manifestos over the past 25 years to see if the spatial representation is the same in both languages or if there are fundamental and significant differences between them.