2.1. Introduction

A methodological way of synthesising Catalan language use in the various areas of the fuel sector is to construct, apply and interpret a value or "index" that draws on the different uses of Catalan in the various aspects studied. Clearly, an index needs to be easy to apply and interpret, otherwise it would not fulfil its raison d'être. Our research into the service station sector and its language uses was the first of its kind. This meant that we had to create all of the protocols, carry out the necessary pilot studies and design a comparatively robust analytical tool that could be used again in later studies.

The metric system we needed to develop had to be easy to apply and sound in its measurements. Each service station could have been evaluated using the traditional method on a scale of one to ten, as occurs with marks awarded at school. This system, which is similar to the decimal metric system, has the advantage that "we are all" familiar with it. However, it also has a considerable disadvantage in that the typology of service stations is largely heterogeneous. This means that we would then have to construct an entire complex weighting system of the indicators making up the index. Even so, the use of compensating systems in an evaluation system would still be questionable. This led us to opt for noting down the characteristics of the variables under study, rather than resorting to scales of evaluation. Our observations could then be summarised using a system that was capable of making statistical comparisons and contrasts.

One alternative to the metric system could well be a probabilistic system, capable of recognising the absence/presence of a fact: the presence of Catalan in service stations. Using this approach, a detailed observation questionnaire was created to note down which language was present in a certain stimulus (either spoken or by way of a written sign). A note was made of the language used in signs, notices and/or posters, whether these were in Catalan, Spanish, bilingual, ambiguous – in the numerous cases in which linguistic ambiguity is used – or the pictogram, when this strategy was used. In the spontaneous interaction section, we also noted the language in which the person working at the service station addressed customers.

This system of collating information allows us to make a more accurate analysis later on, since the reality is often blurred and the categories of the linguistic form in which staff dispense petrol or of each multinational petrol company are not necessarily exclusive. Using this method, we can calculate the probability of finding a particular sign in Catalan at a service station in the Principality (with all the possible combinations).

In order to ensure that the index is a useful tool of measurement, a list was made of all the signs that should or may be present in a petrol station. The prior observation of a small sample may lead to error since some stations did not display the signs required by law. As part of our preparation for the research, we studied a number of designs of stations and the current legislation on signage. Since the size of service stations and their location in the road network could influence the amount of signage at the station and their impact on the population, these were treated as independent variables in the analysis with the aim of contrasting this possible influence using the index of use.

Finally, before suggesting that signage has a possible "subliminal" influence on consumers from a linguistic point of view, we collated further information on the presence/absence of linguistic fact. If we add up the number of signs in a service station where Catalan is present and divide these by the total number of signs observed in all of the petrol stations, this will give us an indicator: the probability of finding texts written in Catalan in that particular station. By analysing the distribution of this indicator throughout Catalonia, we obtain the probability of finding a sign in Catalan at petrol stations in the Principality (with the sampling error calculated as more/less). This methodological strategy can also be applied to other uses of Catalan in service stations, such as the other cases we observed: spontaneous oral use and written use in internal documentation.

This system of collating information was based on a questionnaire carried out on a systematic observation route to evaluate all service stations in the same way. What would happen then with smaller petrol stations? The size of a service station is related to the number of pumps it has and these, in turn, are often related to the average number of customers. Thus, if this data is collected as additional data, it could be used to correct certain deviations. Another system of correction we used was to assess the stimuli found in a significant number of petrol stations in the index. This was one way of indicating that they existed in one language or another or in pictographic form.

2.2. Presentation of Three Indexes of Use

The fuel sector is extremely diverse as a result of the structure of service stations in terms of brand, location and size, in addition to heterogeneity in the ways of selling and market strategies. These can have a significant effect on language use and the processes that produce a lack of standardisation of Catalan throughout the sector. The results of the aforementioned descriptive analyses for each of the areas provides the basis for synthetic study of Catalan use.

In order to systematize and synthesise the linguistic use of Catalan in this sector, we created an index of Catalan use in signage. We used this index to measure the presence of Catalan in a number of signs and notices; we will call this index 1, or index of Catalan use in signage.