136 Anna-Brita Stenström
- Conclusions
This study started out from four research questions. The first question concerned
what functions vale and okay have in common. The material showed that all the
functions performed by vale, directive as well as reactive, were also performed by
okay. The second question concerned whether the total frequency of the markers
differs between the two corpora. As expected, vale turned out to be slightly more
frequent than okay. The third question focused on differences from a sociolinguis-
tic point of view between COLT and COLAm. The data showed that the upper
class girls were the more frequent users of vale, and that the upper class boys were
the most frequent users of okay. The concern of the fourth question was whether
the pragmatic markers vale and okay were typically used by teenagers. Not unex-
pectedly, the study showed that this was the case. Interestingly, it turned out that
okay has not really been adopted by the Spanish teenagers in Madrid, despite the
fact that it has become an extremely frequent international item that has spread all
over the world. There were only four examples altogether, which seems to indicate
that there is no need for okay, since vale does the job.
Clearly, access to corpora for comparative studies is a very great advantage.
But many problems remain. The main problem is that few corpora are collected
along exactly the same lines, which means that they are seldom directly compa-
rable. When it comes to corpora of spontaneous speech, the reason is, first of all,
that it is difficult to achieve the same composition of speakers in terms of gender,
age and sociolinguistic background and, second, that it is impossible to control
what is going on, if the recording is handed over to students with instructions
to record as much as possible for a couple of days and to write down informa-
tion about students and settings in a specific logbook, but without any kind of
control, as in the case of COLT and COLAm. One effect of this is that there is no
guarantee that the students who volunteered to record their conversations with
their friends kept a record of who they were talking to and in what settings. What
one can compare is then rather restricted, and few really safe conclusions can be
drawn. Consequently, there is a need for more reliable comparable corpora of
spoken language and corpora representing more languages to facilitate contras-
tive studies on a large scale.