2. Technical features and methodology

2.1 Computer application

There is an easy to use and clearly useful computer programme which can be utilised both for the organisation of observations and for the introduction and processing of data, as well as for the drafting of reports based on the exportation of graphs manufactured directly by the programme. This is a 32-bit application for Windows constructed from C++ and Visual Basic, which takes up only 20 Mb of space in the hard disc.

2.2 General aspects

In general, all Ofercat data is derived from data collection using sociolinguistic methods of observation, and in fact four different techniques are combined: random route observation of commercial premises and service sector companies, visits to organisations, telephone calls and other observation.

Depending on the district in question and its territorial setting, samples were selected of the different organisations and random routes taken for the effecting of observations.

2.3 Random sampling

Statistical samples are those constructed by random selection. Strategic samples are obtained when we work in strategic areas or areas with a great concentration of commercial premises. Where necessary, both methods were used in combination. With Ofercat, in the case of the samples of commercial premises and service sector companies, we worked with a margin of error of + 5.

To calculate the total number of routes, the number of commercial premises and service sector companies in the sample for each area were divided, approximately, by 50 shops and companies per route. Along these routes we observed identification labelling and informative labelling in the shops and premises, the advertising in the street, and the signs.

The number of routes was selected in the different areas of the town or city in question proportional to the number of commercial premises per area, where we had this information, and in an even distribution where we lacked such information. Specific routes were also selected going to the shopping centres and the main covered markets.

2.4 Visits to organisations

In the organisations visited we observed the two types of labelling: identificational and informative. In the case of identificational notices, the main notice identifying each office or premises was selected. In the case of information notices, the first three notices of this type that were seen were included in the sample.

The number of organisations (institutions, bodies and boards, associations and other organisations) included in the sample, was determined in accordance with table 1.

Table 1. Number of organisations observed

2.5 Telephone calls

The initial language of identification and the language of the subsequent response from the interlocutor at the organisation were noted.

Initial language: the language used by the person answering the telephone before asked any question is made to them. To avoid ambivalent cases ("Si?" ("Yes"), for example, which is the same in Spanish as in Catalan) or in the cases where nothing is said, the caller would remain silent, to force the person to say something. Where nothing is said, the telephone call was null and void, at least as far as the initial response question was concerned.

Language of Response: This is the language used by the interlocutor in response to the question asked of them in Catalan.

Two types of telephone call were made: those made to the sample of shops and service companies, selected at random from the Yellow Pages, and those made to the other organisations. The first were gathered according to the grid for telephone calls to shops and service sector companies. In the second cases observers were supplied with a list featuring the telephone number of the organisations to be telephoned (in most cases, these were the same organisations that we visited personally, with the exception of the schools,which were not telephoned.

2.6 Other observation

Other data was collected by other than the above-mentioned methods. These included:

- Media with the greatest print run or circulation of the various organisations

- The language of the websites

- Specific random routes: factory estates and shopping centres

- Written communications sent to the service industries' customers - water, electricity, gas and telephone.

- Tickets issued by public transport

- Advertisements: shops and surveys, and adventure playgrounds

- Church services and ceremonies

- Local press (magazines, television and radio)

- Language of education

2.7 Further information already included in the programme

These documents, from centralised organisations, which are the same throughout Catalan territory, were collected via the Ofercat Commission and, therefore, their results are included directly in the computer programme.

3. Main results

At the present time we have the results of the pilot Ofercat trial carried out in Santa Coloma de Gramenet [suburb with many residents originally from outside Catalonia], and the first cycle of observations, carried out in the towns and cities of Tremp, Manresa, Tarragona and Lloret de Mar, and the Old City district (Ciutat Vella) in the centre of Barcelona, during 2001.

Consequently, the data we offer refer to the five cities and the district of Ciutat Vella was mostly obtained in 2001, with the sole exception of Santa Coloma de Gramenet, which date from 1998. In spite of this time lapse, we preferred retaining the Santa Coloma data, given that it is the only city in the inner ring of satellite towns for which we have data at this moment.

At the same time it should be noted that in 1998, in Santa Coloma, not as many observations were included as in the standard version of Ofercat, and accordingly this result has to be taken as provisional. However, the information we now have, gathered in 2003, of the same town, which now includes a sufficient number of observations, indicates that the index we have for Santa Coloma dating from 1998 constitutes a good approximation.

First of all we will present the general indices, and move on to the results of each of the subsectors and, lastly, we will present averaged figures for the main subsectors and ambits, and for the district of Santa Coloma de Gramenet by factors.

As the graph shows, the availability of Catalan goes from a maximum value of 78 (in the case of Tremp and Manresa), to a minimum of 46 (in the case of the Barcelona suburb Santa Coloma de Gramenet). What we have here, therefore, is a rather marked step or range of 32, seen in the distance that exists in terms of oral and written competence in Catalan, between one place and another, according to the 1996 linguistic census. Thus, between the 89.36% of the population who claim to speak Catalan in Tremp, and the 52.2% who claim to do so in Santa Coloma de Gramenet, there is gap of 37.16 %; and between the 55.71% in the Manresa sample who said they knew how to write it, and the 32.61% once again in Santa Coloma, there is 23.10 % difference.

Even so, the gradient is not even, but rather hints at different taxonomies, all related to the Catalan availability factor. Clearly, the number of observations or sightings taken to date is still very limited, and it is not possible to draw definite conclusions. We need to wait, therefore, to get the results of the twenty-five towns and cities in the overall Ofercat cycle. Nonetheless, what we have here is a group of populations where the availability of Catalan is majority level (Tremp and Manresa), a group which is quite homogenous; an intermediate group (Tarragona, Lloret and Ciutat Vella), definitely more disperse; and a final downturn, to a "group" where we just have, at present, the results for Santa Coloma de Gramenet.

Figure 2. Ofercat results by sectors

But it is necessary to go into finer detail, and observe the data by sectors as shown in figure 2.

If we calculate tendencies, the sectors which, in most cases, have highest values (score highest) are Public Administration and Education and Health; there is a second block with Leisure and Society, and Economy; and lastly the Media. The only exception to this, in this scheme, is that in Santa Coloma de Gramenet the Economy index or score is lower than the Media one, although, as we indicated above, we have to some extent to take these data as provisional.