Speaking for others: The pros and cons of group advocates using collectives language

Matthew J. Hornsey, Leda Blackwood, Anne O'Brien

Research output: Contribution to journalArticlepeer-review

16 Citations (SciVal)


We examined how rhetorical style affects evaluations of group advocates, and how these evaluations are moderated by group identification. University students were given a letter to the editor defending student welfare. The argument was either constructed using personal language ('I believe') or collective language ('we believe'). Furthermore, the letter was either attributed to an official advocate (president of the student union) or an unofficial advocate (a rank-and-file member of the student body). Consistent with the social identity perspective, participants who showed strong identification as a university student thought that the group would feel better represented by official advocates using collective rather than personal language. Low identifiers, however, did not rate the rhetorical styles differently on representativeness. Furthermore, low identifiers (but not high identifiers) rated official advocates as more likable and more effective when they used personal rather than collective language. The discussion focuses on the conflict low identifiers might feel between (a) needing to homogenize with other group members in order to maximize the influence and political effectiveness of their message at the collective level, and (b) protecting themselves against categorization threat.

Original languageEnglish
Pages (from-to)245-257
Number of pages13
JournalGroup Processes and Intergroup Relations
Issue number3 SPEC. ISS.
Publication statusPublished - Jul 2005


  • Categorization threat
  • Collective action
  • Group advocates
  • Language and intergroup relations

ASJC Scopus subject areas

  • Communication
  • Cultural Studies


Dive into the research topics of 'Speaking for others: The pros and cons of group advocates using collectives language'. Together they form a unique fingerprint.

Cite this