166 Malcah Yaeger-Dror, Tania Granadillo, Shoji Takano, and Lauren Hall-Lew
20 Other situations had also been analyzed in previous English and Japanese corpus
studies and the results are of interest for comparison: US political “discussion”
programs (such as Cross¿ re, MacLaughlin Group) have even higher NEG promi-
nence percentages than political debates which have been analyzed (Kennedy/
Nixon, Bush/Carter, Bush/Clinton/Perot—see further discussion in Yaeger-Dror
et al. 2003)—(78%>55%)—and both are signi¿ cantly more likely to use promi-
nent NEG in remedial turns than nonadversarial conversations (Yaeger-Dror et al.
2003, Hedberg and Yaeger-Dror 2008, Takano 2008). Face-to-face interactions
are not less likely to use prominent NEG than telephone interactions in Japanese
(33%~29%-Takano 2008), but in most English Face to Face conversations studied
the there is less prominence (Yaeger-Dror 1985)^ (3%<31%), even in face to face
group therapy sessions there’s less prominence (Yaeger-Dror 1985)^ (13%<31%),
phone conversations between strangers (such as the Switchboard corpus (Yaeger-
Dror et al. 2003)—13%<31%) or with immediate family members (as in the Call-
Home corpus analyzed by Banuazizi 2003—13%<31%); these are all signi¿ cantly
less likely to use prominent NEG than the CallFriend calls studied here (Yaeger-
Dror et al. 2003; Banuazizi 2003), as shown in the following table.
Overall prominence percentages of NEG in different corpora of English ana-
lyzed to date. Note that the News tokens are all informative, but in conversation
the percentages are for remedial tokens.
Corpus % Reference
Hirschberg’s BUR News 97 Hirschberg 1990, 1993
LDC News 78 Present paper
Political Panel Discussions 78 Hedberg and Yaeger-Dror 2008
Presidential debates 49–65 Yaeger-Dror et al. 2003
Group therapy session 13.3 Yaeger-Dror 1985
SWB 13 Yaeger-Dror et al. 2003
CH 13 Banuazizi 2003
CF 31 Present paper
Face to Face 2.5 Yaeger-Dror 1985, 200221 Note again that all Japanese men were from the Kanto (easter n Japa n) region,
here marked “Tok” for Tokyo, and all women were from Hokkaido (3) or the
Kansai (Western Japan) region (1). The 8 Spanish women are divided evenly
between Costeño and Serrano, and the men were also almost evenly divided.
Note that there were no Southern US English news readers, in our sample.
Within the US North, the regions were roughly divided into West (=W), Inland
North (=nc), East (=E), and Ashkenazy (=y), while Southern speakers, based on
Feagin’s work and the ANAE, are divided into those from formerly “rless” areas
(=S) and those from fully “rful” areas (=A).