How did the term “liberal” become a dirty word in American politics? Why are Democrats “accused” of being liberal as a campaign tactic? These types of questions motivate Sharon Jarvis's book. The language of politics, especially partisan politics, is a topic visited by political scientists and other social scientists with modest frequency, but Jarvis offers a different, and fresh, perspective that bridges public opinion approaches with rhetorical analysis. This method should prove appealing to a variety of social scientists and students and illuminate new ways of studying political language.

Jarvis makes four assumptions that also serve as themes for The Talk of the Party: “[First,] citizens come to know politics through discourse[;] second, citizen discourse is led, but not determined, by elite discourse[;] third, the uses of labels function as powerful shortcuts in modern life … and fourth, the meanings of political terms shift with time” (p. 3). To students of public opinion, Jarvis's framework for analyzing the branding of political parties is a familiar and useful one. This theoretical foundation is laid out in the first two chapters of the book. The next four chapters take these themes and emphasize the role of elites in shaping the perceptions of the masses with regard to the party labels, symbols, and brand names. In the final chapter, the author returns to questions that motivated the book and offers her explanation, given the preceding analysis, of how we come to see the uses of political language and interpretation that are evident in recent campaigns.

Jarvis employs an extensive content analysis as her methodology. She relies on discourse from campaigns (from journalists and presidential candidates), government (debates in Congress), and education (civics textbooks). The analysis covers five decades, beginning in 1948. She focuses her content analysis on six key terms: “Democrat,” “Republican,” “Independent,” “party,” “liberal,” and “conservative.” This approach allows for an in-depth treatment of the discourse surrounding the terms and their uses, and a longitudinal analysis to detect change. Implicit in this design are assumptions about the role of certain elites in “managing” the discourse of politics, or at least these terms, and the extensive reach of that discourse to the mass public.

In addition, and particularly impressive, is a valuable methodological appendix. Jarvis provides a comprehensive description of the content analysis, from sample selection to coding to interpretive tables of data. The decision making is clear and reasonable. The appendix is an excellent example of how to “show one's work” and should prove a useful resource to scholars and students alike.

One of the most insightful of Jarvis's observations is that party discourse is hierarchical. She demonstrates that political parties are mentioned, described, and associated with political elites. As she notes, this “elite nature” of party labels provides an interesting way to deal with the seeming paradox presented by scholars and others about why Americans hate politics and politicians when they are so patriotic (e.g., see E. J. Dionne, Why Americans Hate Politics, 1991). As the author explains in Chapter 3 and details later, “As long as these voices continue to regard party labels as elites do, then party labels' role in government is, in a semantic sense, secure, even if occasionally vilified” (p. 80).

In Chapter 4, Jarvis claims that four periods of party representations emerged from the data she collected: monolithic, liminal, fragmented, and resurgent. The usefulness of these terms to describe the over-time changes in how parties are portrayed in the media and how party leaders portray themselves is difficult, at best, to appreciate. Her labels lack identifiable and portable features. What constitutes a “fragmented” state of parties, for example? The list of societal characteristics and technological developments do not define a fragmented or any other party or period of party politics. Likewise, while the data show a high level of “party language” in the period between 1948 and 1960, why that constitutes a monolithic period is unclear. An alternative, and not new, interpretation of the data in Figure 4.1, for example, is that they support the observation that parties were strong in the late 1940s (and probably in the decade prior), gradually declined into the 1980s, and have recently transitioned (resurged?) to better fit the needs of modern politics (i.e., becoming conduits for campaign funds, providing campaign resources for candidates, etc.).

It is also tempting to conclude that it is difficult to discern a difference between Eras II and III (liminal and fragmented party periods), for example, because that difference appears to be largely in the eye of the reader. The departure from the statistical analysis and reliance on examples from presidential campaign speeches and journalistic accounts proves a much less persuasive tack. Evidence to show the usefulness of demarcating periods in the way suggested is not marshaled.

In the fifth chapter, Jarvis examines how different types of elites view political parties and how they portray them to others. This is, perhaps, the most interesting of the chapters in the book. When we read or hear legislators, journalists, educators, and presidential candidates communicate about political parties, they do so with varying descriptors and tones. In the process, the meanings they attach to the labels are a cacophony of messages to the mass public. Jarvis shows not only that the labels elites use to talk about parties change in frequency and meaning over time, a reasonable expectation, but that even at any one historical period, they do not share the same rhetoric. The implications are intriguing as much for political scientists as they are for the mass public. Stated well by the author, “political scientists feed on a richer diet of party meanings than do citizens” (p. 154).

One minor shortfall of the book is that a significant number of new concepts and variables in the author's analysis are introduced with little explanation as to their meaning and measurement. While the methodological appendix is excellent, this is one of those cases where more of the type of information contained there should be placed in the text. Terms such as “quality,” “context,” and “task,” to name a few, are integral parts of the analysis but require a trip to the back of the book.

For political scientists versed in social psychological theories of public opinion formation (e.g., political socialization, opinion formation and change, etc.), Jarvis's contribution lies in providing views of elite discourse in real terms. For students of political parties, this book offers a rich store of detail about the way parties are portrayed by the managers of political discourse and how those portrayals have changed over time. Scholars of discourse, political or otherwise, will find an excellent example of a well-executed analysis of political text and the process of branding. For both groups, this book is useful as a supplement for undergraduates in courses that emphasize understanding and appreciating the importance of symbols in American political life.