In this comment, I argue that ideology should be examined via looking at the structure of how different political beliefs and attitudes are related with each other, instead of looking at how political issues are phrased and expressed, or, discursive styles.
This article, “What Drives Media Slant?", is based on the assumption that ideological properties of a newspaper can be understood by its discursive styles. Drawing on such theoretical assumption, this article elicits ideological phrases (i.e politically charged phrases) from congressmen's speech, compares the discursive structure of politicians and newspapers, and measure the degree of a newspaper's ideology.
Such theoretical assumption is, in turn, based on Geertzian understanding of ideology. Clifford Geertz understands ideology as “schematic images of social order” (Geertz, Interpretation of Culture, 218), which provides symbolic frames, a “a compelling image of social reality” (Geertz, 229) and helps individuals navigate through a myriad of unfamiliar events. Just as this article focuses on the patterns of how newspapers 'name' and 'phrase' political issues, (such as 'death tax' and 'estate tax'), Geertz focuses on how political issues are phrased in terms of “tropes” or “metaphor” (Geertz, 210). (Geertz's example here is "slave labor act"). Seen this way, ideological properties are embedded in the ways in which discourses are expressed.
In a nutshell, this article has the following theoretical assumption: first, the ways in which political issues are expressed and phrased imply the ideological properties (if someone uses 'death tax', then he/she understands this issue from conservative ideological perspective); and thus, second, ideological properties can be measured by the extent of discursive similarity.
It is true that discursive structures are directly related with, or causally linked with, ideology, but such approach fails to look at the essential conceptual attributes of ideology. Ideology is a set of beliefs shared by a group of people (Gerring 1997; Converse 1964), and thus a proper measurement of it should measure the degree of consistency in the belief structure, instead of looking at how political issues are phrased and expressed. It is very likely that an ideological person would show ideological phrases of the her ideological group, but such relation between ideology and ideological discursive characters should be investigated, instead of being assumed.
To wit, an entity is ideological when one's political beliefs and positions are related with each other in a principled, shared way. A person is ideological when one's belief system is constituted by one political principle. For example, a person is ideological when one's belief system is completely liberal in every dimension--pro-abortion, social liberal, pro-government intervention and so on. Thus, ideological properties should be measured by the degree to which an entity's political positions along various dimensions are related with each other. Ideology should be measured by the degree of "constraint" (Converse 1964; Martin 2001; Della Posta et al 2015).
I think we can actually measure the ideologicality of newspaper by looking at the patterns of political viewpoints via text-analysis. We can find the discursive structure of the left and right, or Democrats and Republicans. Drawing on such techniques, we can look at the political positions of each newspaper along various political dimensions, and, thus, we can see how each newspaper organizes their political viewpoints. This is what Hans Noel (2014) did to examine ideological properties in the media, but he didn't employ text-analysis. (He hand-coded every piece...)
In this comment, I argue that ideology should be examined via looking at the structure of how different political beliefs and attitudes are related with each other, instead of looking at how political issues are phrased and expressed, or, discursive styles.
This article, “What Drives Media Slant?", is based on the assumption that ideological properties of a newspaper can be understood by its discursive styles. Drawing on such theoretical assumption, this article elicits ideological phrases (i.e politically charged phrases) from congressmen's speech, compares the discursive structure of politicians and newspapers, and measure the degree of a newspaper's ideology.
Such theoretical assumption is, in turn, based on Geertzian understanding of ideology. Clifford Geertz understands ideology as “schematic images of social order” (Geertz, Interpretation of Culture, 218), which provides symbolic frames, a “a compelling image of social reality” (Geertz, 229) and helps individuals navigate through a myriad of unfamiliar events. Just as this article focuses on the patterns of how newspapers 'name' and 'phrase' political issues, (such as 'death tax' and 'estate tax'), Geertz focuses on how political issues are phrased in terms of “tropes” or “metaphor” (Geertz, 210). (Geertz's example here is "slave labor act"). Seen this way, ideological properties are embedded in the ways in which discourses are expressed.
In a nutshell, this article has the following theoretical assumption: first, the ways in which political issues are expressed and phrased imply the ideological properties (if someone uses 'death tax', then he/she understands this issue from conservative ideological perspective); and thus, second, ideological properties can be measured by the extent of discursive similarity.
It is true that discursive structures are directly related with, or causally linked with, ideology, but such approach fails to look at the essential conceptual attributes of ideology. Ideology is a set of beliefs shared by a group of people (Gerring 1997; Converse 1964), and thus a proper measurement of it should measure the degree of consistency in the belief structure, instead of looking at how political issues are phrased and expressed. It is very likely that an ideological person would show ideological phrases of the her ideological group, but such relation between ideology and ideological discursive characters should be investigated, instead of being assumed.
To wit, an entity is ideological when one's political beliefs and positions are related with each other in a principled, shared way. A person is ideological when one's belief system is constituted by one political principle. For example, a person is ideological when one's belief system is completely liberal in every dimension--pro-abortion, social liberal, pro-government intervention and so on. Thus, ideological properties should be measured by the degree to which an entity's political positions along various dimensions are related with each other. Ideology should be measured by the degree of "constraint" (Converse 1964; Martin 2001; Della Posta et al 2015).
I think we can actually measure the ideologicality of newspaper by looking at the patterns of political viewpoints via text-analysis. We can find the discursive structure of the left and right, or Democrats and Republicans. Drawing on such techniques, we can look at the political positions of each newspaper along various political dimensions, and, thus, we can see how each newspaper organizes their political viewpoints. This is what Hans Noel (2014) did to examine ideological properties in the media, but he didn't employ text-analysis. (He hand-coded every piece...)