102 LANGUAGE AND THE INTERNET
greetings defies easy generalization. Other factors than social rela-
tionship enter in: only a mixture of subject-matter, time-pressure,
and mood can explain why my editor at Cambridge University
Press switches throughout the year fromDavidtoDear David(in
a ratio of 1:2) in his messages, and doubtless I am just as variable
in my labelling of others.^11
Another factor is the location of the name, once it is used. The
majority of my messages place the greeting at the head of the
message body, usually spaced away from the maintext as in a tra-
ditional letter. This is always the case in +Dearopenings. With
informal –Dearopenings, however, the location varies. It is most
often spaced and separate (in a ratio of 3:1,in my corpus). When
it is on the same line, it is usually the first word, but is sometimes
placed later, especially in replies (Thanks,David;OKDavid;Thanks
for your message, David), which to my intuition is more informal
than an initial placement. It is unusual for an inserted name to ap-
pear much later in the opening paragraph, or in later paragraphs –
though occasionally one finds instances of ‘rapport renewal’, such
as (from a third paragraph):
Sorry to put you to this bother, David, but...This is no different from what is done in traditional informal letter-
writing.
Farewells display fewer possibilities for variation, but the same
points of principle arise. Two elements are available: a pre-closing
formula (of theBest wishestype) and the identification (ID) of
the sender. Most interpersonal messages (80%, in my case) end
with both elements present, and the influence of traditional letter-
writing is evident in the overwhelming tendency to place each
element on a separate line, usually spaced away from the mes-
sage body. The remaining 20% give a name, and dispense with the
(^11) In Gains (1998), only 9 out of 54 usedDearinterpersonally, and only 1 out of 62 institu-
tionally. By contrast, Li Lan (2000) found 31 out of 77 usingDearinterpersonally and 35
out of 76 institutionally – again suggestive of the existence of a pull towards traditional
usage in non-native speaker settings.