One of the main interests behind this study has been to examine how institutions, such as the social services, deal with people 's troubles and difficulties, how a private problem is transferred to the language used within a public institution. The study deals with the processing of applications for financial assistance, and is based on a documentation of talk and text - from the initial telephone mediated contact to the final written decision - which are produced in the handling of the clients' applications. Even though previous studies of institutional discourse have shown interest in the daily routines of formal organizations, working with the processing of cases, they have often neglected the role that different kinds of texts play in these processes. The present study concems the roJe of text and talk as parts of such routines.
DATA
The study contains a detailed documentation of the processing of appJications for financial assistance involving 25 individual cases. This corpus was compiled <luring 1988 and 1989 in three social welfare offices in two middle-sized Swedish cities. It comprises mainly the following data:
a) telephone conversations between social workers and clients (20 calls are documented by means of participant observations and written notes)
b) face-to-face conversations between social workers and clients (25 audio-taped conversations)
c) application forms and other written documents which form the basis for decisions (copies of the relevant papers from 20 cases)
d) the social workers' written decisions (copies from 20 cases)
e) meetings with the social welfare committee (discuss ions and decisions conceming 3 applications documented by means of participant observation and written notes: copies of the minutes and other kinds of written decisions)
f) interviews with the clients and social workers, who participatedin the study (23 clients and 15 social workers have participated in audio-taped interviews).
ANALYSIS
The recorded conversations and interviews were transcribed in extenso. The analyses are based on the transcriptions as well as the audio-tapes. The conversations have been analysed in several ways, e.g. in terms of sequential structure and contents as well as their function in the institutional context.
The written documents - the applications and decisions etc. - included in the study have been analysed mainly with regard to their significance in the processing of the individual cases. The documents have made it possible to investigate both the relation between spoken and written discourse and the processing of 'cases' through several stages, which is common in many areas of bureaucratic organizations (Jönsson, 1988).
RESULTS
The presentation of the results of this study follows the stages which have been identified in the processing of the cases. This handling of the clients is found to follow a distinct order which includes the following steps: 1) the client's initial telephone call to the social welfare office; 2) the application form which is supposed to be filled in by the client; 3) the social welfare interview; 4) the decision concerning the client's application. In short, the findings may be summarized as follows.
The initial telephone conversation functions as a buffer between the social welfare officer and the applicants. During this stage in the processing of people, a 'screening' takes place. This gives the social worker a chance to make informal judgements as to whether the client should persist in seeking help and also to provide the ' client-to-be' with information concerning other ways of solving her or his financial difficulties. If the client's financial (and social) position seems to fit into some of the categories that make a person eligible for financial assistance, a slot for the client is found (i.e. this person is offered an appointment for an interview and/or provided with an application form).
The application form is used to collect information about the applicant. This form, however, plays a more prominent role in the different stages than expected, since it gives a structure to many of the routines and procedures used when a client's application is being handled. This procedure may moreover be seen as a way of disciplining clients, and teaching them 'the client role' (cf. Lipsky, 1980, p. 61). When a person has filled in the necessary details on this application form, he or she has been accepted and given a client status in the further relations with the social welfare office.
The social welfare interview gives the client access to the social welfare office and gives him the opportunity of providing more information about his or her financial difficulties. The conversation usually begins with a revision of the application form. In connection with this, the client often talks about the events in the past which are related to the need for financial assistance (cf. Cedersund, 1992). It is during the interview that we are able to observe how the client's financial problems are reformulated in a way that may be accepted within the social welfare organization. This transformation of 'problems' into 'cases' is made possible through the use of a typification where the clients are assigned to certain categories which are usually implicitly expressed in discourse.
The decision is made by the social workers or in some cases her supervisor and/or the district board for social welfare services (Sw. : social distriktsnämnd). The decision is often given to the client in the social welfare interviews and the written decisions are mostly produced after this encounter. These written texts in most cases include a short summary of the specific details of the client's financial difficulties and needs. This summary may be similar to a later version of the client's own oral account of his or her financial difficulties given during the welfare interview. The last part of the written decisions always includes statements which show the approval (or rejection) of the client's application. The written decisions seem to be produced in order to document the processing of the applications within the office rather than as a text to be read by the clients.