Phraseology and Culture in English

4. Conversational routines and frames 
All language is embedded in a social and cultural context. When we look at 
routinized phrases it is clear that the dependence on context goes deep. 
What makes routine phrases special is that the links to features of the social 
and cultural context are conventionalized. 
Routine phrases in answering machine messages must be associated 
with a rich description of the context. We need to take into account ‘who 
speaks what, why and under what circumstances’ (cf. Fishman 1970: 364) 
in order to describe the context. The context also includes such factors as 
the user’s attitude to the medium and knowledge about the medium and 
practice in using it. Although the cultural context seems to be of interest it 
has not been shown that the message form varies depending on culture. 
Dingwall (1995) collected receivers’ messages from the German part of 
Switzerland, which she compared with messages in French and Romansch, 
also from Switzerland. She found no cultural differences between the mes-
sages in different languages which were recorded under the same condi-
tions although there were some differences in style between the German 
and the French material. 
The features of the social context can be organized as (social or situ-
ational) frames. Fillmore’s notion ‘frame’ (1985) and Levinson’s ‘activity 
type’ (1979) consider context in a broad sense and can be used as a model 
when one wants to describe how routine phrases are linked to particular 
speech events and to context. A frame is defined ‘as a set of lexical items 

---

328
Karin Aijmer 
whose members index portions of some actional or conceptual whole’ (Hanks 
1996: 243). 
Hanks (1996) discusses how frames play a role for the description of a 
rich ‘ethnographic context’ needed for the interpretation of deictic systems. 
The frame can be conceived of as a list of strategically ordered questions 
(Who are participants in the event? In what setting does the event take 
place? For what reason? What is the preceding and subsequent speech act? 
How is it performed?). Speech acts often have a conventionalized form and 
are described as frames. The following frame (from Coulmas 1979) illus-
trates the factors needed to describe the meaning of ‘congratulations’: 
I Participants 
sex 
age 
social 
role 
hierarchy 
authority 
familiarity 
II Setting 
time 
place 
III 
The why and wherefore 
time 
reason 
IV Contextual 
restrictions 
sequentialization 
stylistic 
homogeneity 
V Concomitant 
activity 
e.g. 
gesture 
The situational frame plays an important part in describing the associa-
tion between form and situation which goes beyond the conventional link 
between form and an illocutionary force indicating device (cf. Searle 1969: 
30) or generalised conversational implicatures as in Levinson’s theory (cf. 
Section 6). 
Fillmore (1985) was mainly interested in how frames could be used in 
lexical semantics for instance to establish lexical fields. The key idea is that 
we can only know the meaning of words in a frame by considering other 
words in the frame. For example, we can only understand the meaning of 

---

Idiomaticity in a cultural and activity type perspective
329
buy and sell by considering the commercial transaction of which they are 
part.
There are also frames for telephone conversation, and many other events 
or transactions characterised by a sequential ordering of events. Kiefer (un-
published) discusses the association between ‘bound utterances’ and ‘sub-
events of a frame’. A commercial transaction consists for instance of a num-
ber of subevents which may or must be accompanied by one or several 
bound utterances. I will also discuss other features which are part of the 
meaning or frames of conversational routines for example setting (institu-
tionalised), social roles (Do the caller and answerer know each other?), goal 
(asking the other person to call later, asking for information, etc.). 
4.1. Frames for answering machine messages 
A frame (script or ‘blueprint’) for an answering machine message might 
look as follows. It consists of ordered (sub)events such as ‘opening’ (what
happens first), the ‘message topic’ (what is the goal of the communica-
tion?), ‘closing’ (how do we know that the message is closed?). Other fac-
tors have to do with attitudes to the medium (conceptual restrictions). If 
people view the answering machine only as a machine they will use differ-
ent strategies than if they address the other person as if he or she were actu-
ally present on the communication channel. For example a greeting phrase 
suggests that you are aware of addressing a person rather than a machine. 
Another factor has to do with whether people view the medium as spoken 
or written. For example, some messages are similar to letters in the way 
they end. 
As shown by the example, a particular event may, but need not be ac-
companied by a routinized phrase (Kiefer 1996, ‘bound utterance’). The 
frames on the answering machine have a special form depending on whether 
the fact that a greeting is not reciprocated, the identification is not acknowl-
edged and the request is ‘left in the air’. There are several different message 
forms as illustrated by messages from the two corpora: 
A. Greeting + other identification: 
Good morning Annie 
Caller’s identification: 
This is John Marshall 
Message: Could 
you 
please phone me at X 
Closing: 
OK thank you bye
(Surrey) 

---

330
Karin Aijmer 
Some frame features: 
Medium 
the answering machine 
Setting
university department 
end of 90´s 
Participants
caller (identity unknown) 
The caller is responding to a message left by a person of whom they 
know the identity of but not necessarily someone they actually know. 
The why and wherefore 
asking someone to call back 
Contextual restrictions 
sequentialisation
Style
formal 
Conceptual restrictions 
The caller views leaving messages on the answering machine as com-
municating with a machine or with a real person. 
An ‘event’ such as opening (cf. Kiefer 1996) can be broken up into the sub-
events greeting and identification. The events (and subevents) are sequen-
tially ordered. The caller’s identification comes after the greeting and be-
fore the message (a request to call back). There are routines for greeting, 
self-identification, asking the other person to call up and terminating the 
message. 
Frames and the routines which go into them are largely routinized. 
However they may be varied if this fulfils a purpose: 
B. Greeting + answerer identification: 
hi Peter 
Caller’s identification: 
Henry 
Message topic: 
been some time
Closing: ta 
(Surrey) 
When the greeting is followed by a name (hi Peter), naming identifies the 
answerer. In the next variant answerer identification is not preceded by a 
greeting:

---

Idiomaticity in a cultural and activity type perspective
331
C. Answerer identification: 
Olive 
Caller’s identification: 
Julian 
Message topic: 
I am here until you want to ring me 
Closing: OK 
(Surrey) 
The message, which is elliptic, could be spelled out as: (Hello) Olive. This 
is Julian. I am here until you want to ring me. 
D. Answerer identification: 
Maxwell 
Caller’s identification: 
Tony 
Message topic: 
could you phone 
(Surrey) 
To sum up, messages on the answering machine seem to occur in a fixed 
sequence although there is some structural variation. The particular frame 
used to leave messages on the answering machine is structured in a way 
which is typical for the answering machine. Categories which are (almost) 
obligatory on the telephone (such as greetings) can be omitted when one 
leaves a message on the answering machine. The caller’s identification is 
more important on the answering machine, since the answerer cannot verify 
who is calling (cf. Is that Mrs X?) The messages are generally short and 
they often have a fixed content (‘asking the other person to call back’). 
What frequently differs is the form of the routine which is used in the situa-
tion. The elliptical form for self-identification has become routinized (in 
some) messages on the answering machine. 
When the speaker uses phrases instead of whole sentences, the depend-
ence on the context is maximal. The use of elliptical structures will be fur-
ther discussed in Section 5.