An objective study of the language personality phenomenon must refer to all kinds of discursive practices of the individual in their real correlation. However, the materials of informal everyday communication representing the person most reliably are insufficiently involved for their collection is very challenging. The article highlights the Tomsk Linguistic School experience of using effective methods of recording spontaneous speech of certain native speakers. The research of the dialect language personality of a Russia old-timer in Siberia employed the method of inclusion into the language existence of the speaker. Its implementation meant a trusting relationship with the object, regular long-term observation, a comfortable environment for the individual in the process of recording. The peasant woman's speech was recorded in her every-day environment with familiar communicants including her co-villagers, dialect speakers. The result of the long-term project (57 expeditions in 1981-2004) is the archive of 10,000 pages of printed text that reflects the discourse of a typical representative of a traditional dialect. The advantages of the method are its versatility (can be applied to all types of modem language personalities) and direct observation of speech communication by the collector. The disadvantages are its complexity, the need to overcome the negative reaction to the invasion of the individual's private sphere, little informative value of the obtained material for those who did not participate in their collection. The study of language personality belonging to the unwritten vemacular speech culture type is carried out by the individual's self-recording. The use of the new method is conditioned by the lifestyle of vemacular speakers (frequent changes of residence and occupation, working in same-sex men's groups) and by the desire to minimize the impact of the speech of educated interlocutors. This way, the collector is only an intermediary giving records to the university-contractor. The informant was explained that researches had no interest in the content of the speech. It was important, however, to speak in a usual way for a special computer program recognized unnatural texts and they were not considered. The vemacular speaker recorded her speech to an MP-3 player and regularly gave audio material to the researchers. About 60 hours of recorded oral speech were collected in two years. This method does not require a constant presence of the observer, saves the researchers' time and puts less constrains on the informant without breaking her habitual way of life. Self-recording can be successfully used by middle-aged and young people who can use simple modem appliances. Large amount of material can be collected in a relatively short time. At the same time self-recording is not a convenient method for representatives of the traditional folk speech culture, illiterate elderly dialect speakers. The downside is the fact that in this case the linguist does not witness speech contacts of the informant, and the resulting audio material is not always self-contained for adequate interpretation. The informant can choose of what to record, which creates some limitations in comprehension of the language personality in comparison with participant observation. Both methods require informal relations between the linguist and the informant. Researchers also have certain ethical obligations: stealth recording is forbidden, material for the open access must be filtered not to violate the rights and interests of the communicants.
