Since I took over the subject of "Dialectología hispánica" ('Hispanic Dialectology') in the Degree in Hispanic Philology at the Autonomous University of Madrid (in Spanish, Universidad Autónoma de Madrid, UAM) back in 1987-1988, I immersed myself in the then available bibliography on the subject. Several shortcomings soon became evident to me: firstly, the preferential attention devoted to phonetic and lexical dialectology; secondly, the scarcity of monographs focusing on Peninsular Spanish, in contrast to the abundant bibliography of studies on Asturleonese and Aragonese. And, lastly, the absence of textual corpora that would allow me to find answers to the questions that worried me, since the typical dialectal studies did not usually transcribe genuine materials, except for lists of words or, if lucky, a phrase or two. If the monograph presented dialect texts, they were rarely transcriptions of spoken discourse, but rather contrived compositions created either by the usual dialectologist or by some local "authority" on the dialect, who wanted to show in that brief passage its most salient "features". Dialectology focused, moreover, on the detection of non-standard or differential solutions with respect to the standard language, without recognising their simultaneous coexistence with variants of a more general nature. It did not seem possible, in this situation, to undertake an adequate study of many of the dialectal forms or structures of Spanish: the descriptive shortcomings limited the theoretical analysis of the data.

Faced with these shortcomings (which still cannot be considered to have been made up for), I felt that dialectology at the UAM should cease to be a subject based solely on bibliographical sources and that the reality of the Spanish spoken in rural areas should be verified on the ground. I was inspired by the example of Diego Catalán whom I had the good fortune to listen to as a professor of "Dialectología" ('Dialectology'). Catalán introduced a theoretical renewal in the dialectology of his time when he applied structuralism to the interpretation of dialectal solutions back in the 1950s, but also was an expert field researcher whom I was privileged to accompany on several surveys devoted to the collection of the traditional romancero. Encouraged by this aim, I obtained support from the Department of Spanish Philology and the Faculty of Arts to organise a dialectology fieldwork trip, the subsidy for which had to be processed each year until, years later, the activity was recognised by the Rectorate of the UAM as "Prácticas de campo" of the Faculty of Arts. A large part of the COSER recordings have been collected during these fieldwork trips and have been transcribed in preliminary form as part of coursework or final degree projects. Another part has been collected and transcribed in the framework of specific research projects (see Funding).

Initially, the surveys were designed primarily to document the use of clitics, but it soon became clear that it was possible to study many other aspects subject to variation that had not received the attention they deserved. Since then, it has been possible to advance our knowledge of various grammatical aspects whose interpretation had hitherto been conditioned by the paucity of data present in the written language or in the speech of the most educated socio-cultural groups. This advance in our understanding of certain phenomena, such as the peninsular use of unstressed third person object pronouns, the so-called neutro de materia (mass / count distinctions), the shift from the imperfect subjunctive to the indicative and the use of reflexive pronouns, has crystallised in various publications and doctoral theses from the 1990s to the present day. To this interest has been added the possibility of investigating some phonetic changes, especially in their relation to grammar.

The history of COSER is in my opinion an example of how research and teaching can be combined in such a way that both are enriched and benefit from a relatively modest investment. Thanks to the continuous support implied by the teaching activity and the participation of multiple generations of students, it has been possible to compile a corpus whose magnitude would have required an enormous and lengthy investment effort, difficult to reconcile with the usual deadlines for research projects. Furthermore, through the research approach, students are confronted, as part of their learning, with applied dimensions of the subject. On the one hand, they come across the methodology of sociolinguistic interviewing and the problems it raises through real-life experience. On the other hand, the recorded conversation becomes the starting material for the coursework, in which the student is confronted with the "raw" linguistic data, thus enjoying the opportunity to confront them with the specific bibliography and to test his or her capacity for analysis and reasoning on real, rather than purely bookish, material. In addition, the data collected in COSER have made it possible to create an extensive repertoire of dialectal phenomena which are offered as dialectal samples, both to UAM students and to other potential users interested in Spanish dialectology.

There are many other lessons to be learned from the activities around COSER. The collective nature of the process necessarily requires teamwork and the collaboration of the various links in the work chain. Coexistence and teamwork say a lot about people, their qualities and their ability to tolerate, to reconcile, to stick to protocols and to achieve a result. Over the years, I have witnessed friendships and courtships - though I must also confess, also a few conflicts. I have also had the good fortune to enjoy the recurrent fidelity of a few evergreen survey addicts and their inexhaustible curiosity about rural speech.

Alongside the interviewers, the informants have been a constant stimulus to keep on working. The COSER recordings offer numerous personal testimonies, often more interesting in their human than in their linguistic dimension, about life in rural Spain in the second half of the 20th century. When dealing with what life was like "back then", informants tend to exemplify their own or their family's life, which is thus often recounted, albeit not always in chronological order, in a way that is comparable and complementary to oral history projects. We have listened to appalling stories of death and misery, of pain and injustice, but also of joy and nostalgia. We have interviewed people who are talkative and intelligent and others - the fewest - taciturn and obscure. We have learned from all of them.

Inés Fernández-Ordóñez
November 2015