Auto documentation

The Documentation of Tol project is moving forward and the latest development has been my teaching a person from San Juan who is one of the few speakers who writes the language, to use the linguistic annotation program ELAN! This means that he is now able to transcribe, translate and annotate recordings on his own while I continue to make headway with the other speaker who has been working with me all along. This means that we are now 3 team members and more importantly, it means that we will be recording more and annotating more all of which adds value to the growing corpus of spoken Tol. Also, it is my intention that both Federico and Aurelio, who is our new guy on the team, will shortly learn how to operate the video and audio equipment to allow them to go out and make recorings on their own - I think such recordings will have several advantages over ones where I choose what to record and where I am present at the beginning and end of the recording. First of all, the community members know best what might be interesting elements of their culture / community life to document and they have much more easy access to all sorts of situations, and secondly we have the ever nagging observers dilemma where the social scientist sets out to document / observe a culture or language and for that reason needs to be present, but while doing this alters the situation by his mere presence. Of course this cannot be eliminated just by letting community members do the recording, since it is not normal for Tolpan people (or for ANYBODY?!) to go around asking for permission to film each other, but at least it may diminish these distorting effects. It would be great if we could get to a point where Aurelio (who is also one of the only Tol speakers who knows how to operate a PC) could run the entire gamut of documentation activities from recording, file compression, meta data recording, annotation and archiving, however Im afraid the time will not be sufficient for that this time around, and another issue is the lack of funding for future documentary activities once the current project terminates in late March of this year.

Halfway evaluation

All right, so all survey data are now collected, the final 4 interviews having been carried out with two men and two women from the tribe on the other side of the river; this is good news because it is generally assumed to be a difficult task for a single person / project to communicate and work with more than one of these two tribes without becoming unpopular with one of them. This brought the total number of interviewees to 31 and, as it turns out this constitutes 6,6 % of all tol speakers (470, more on this later), and I think we have avoided the worst sampling traps in terms of age, gender and geographical heaping. The upcoming work on the analysis will show whether or not something useful may actually be gleaned from this survey - in any event the recordings in themselves constitute a valuable contribution to the documentation of Tol, particularly in those cases where interviewees would interact at length with the interviewer on topics relating to the questions asked. There are two major issues that have proven difficult to overcome in the interviewing process: First, as I have mentioned before, some of the things that social scientist are interested in knowing have to do with concepts that most of the people we interviewed did not seem to have in their conceptual vocabulary; for instance the concept of paying attention the switch between Tol and Spanish according to social situation / setting does not seem commonplace among the Tol speaking population which mainly consists of farmers who hardly ever leave their communities and almost exclusively sow and harvest for personal consumption; and this in spite of the fact that there are Tolupanes who do not speak Tol and also ladinos living in some of the communities, especially San Juan, the largest one. Secondly, there have been cases where the translation into Tol of my original questions in Spanish have more than one possible interpretation, and so the meaning of the actual question that is asked becomes slightly and in some cases severely altered.

Because Federico, my main collaborator on this project, had the flu for almost 3 weeks before he finally got a serious amount of antibiotic injected into his body and got better in a matter of days, the survey got delayed, and in the mean time I got a brother of his, Ricardo, on board the project and we started recording, transcribing and analyzing spontaneous speech from various contexts. One particularly demanding piece of documentation has been a 2 hours long recording in the front yard of a house with a mother, her 5 children of varying ages, her grandson and various passers by. This has proved difficult to transcribe especially because it was recorded with a single condenser microphone sitting on the ground at an equal distant of about 1,5 meters from most of the participants, meaning that the extremely frequent, and expected, overlaps in speaker contributions to the conversation at times become difficult to separate. In spite of these difficulties it is a beautiful recording of multiple way interaction with the surprising feature of an almost equal number of speaker turns divided between the 3 adult participants: mother, son and daughter; both of my collaborators reacted to this recording with a "wow, that's a beautiful piece of work" (es un trabajo muy bonito).

Wrapping up the survey

Getting the last interviews done has proven more difficult than I first anticipated, mainly due to my criterion of geographic representativity - that is, I wanted to go to the outlying communities to interview people there in order to avoid an overrepresentation of people from San Juan, the main speaker community. Theoretically this is a good criterion because there are many differences between people there and the ones in the more remote communities; for instance, in San Juan there is a higher degree of literacy, although the majority of adult are probably largely ilitterate, there are individuals who have gone to do various forms of work in the capital Tegucigalpa and other cities, and they generally have frequent contact with outsiders, both hondurans (ladinos) and North Americans (gringos). In practice, however, interviewing in the outlying population has proven somewhat difficult - in part due to the mere distances to be walked through mountainous and at times slippery terrain, and partly because in many cases, upon arrival at people's houses, they've either fled to avoid their exotic would be visitor because they have little experience with outsiders and because they are affraid of catching desease from outsiders, or they simply will not participate in an interview even knowing that interviews are conducted in their mother tounge, not Spanish, and knowing that they will receive money in return for their time and effort. Some believe that they have nothing to contribute or that they are somehow inadequately equipped for interacting with us, that is, my speaker assistant from San Juan, Federico, and me. Another issue is the same as we have encountered in San Juan, that women who are at home without their husbands or other men when we arrive at the house will not speak to us without the presence and permission of male family members. For these reasons, and due to two cases of the flu, first me and now Federico, it has taken longer than anticipated to finish the remainder of interviews - we now have 27 and need another 30.

Speaking of which, as a note on sampling methodology, the total number of interviews was settled on partly due to a certain time limit for the interviewing part which could not be allowed to take up more than a minor part of the total project, that is somwhere between 1 and 2 months. The other consideration was that, even though the total number of speakers is still yet unknown (hopefully it will not remain so at the end of this survey) I provisionally assumed a total of betwen 200 and 400 speakers based on earlier reports and an estimate made by Jeff Pynes, my North American colleague in Tol linguistics - so, if these numbers are approximately right, then a sample of 30 speakers with some geographic and demographic (age and gender primarily) spreading would in fact be a 10 % sampling rate which would be considered more than adequately representative by any social scientist.

Once we finish the last 3 interviews, Federico and I are going to go over a list of all Tol speaking communities, probably some 20 locations with between 3 and 50 houses, noting any information that Federico has (or other speakers in San Juan that he may consult with) about the number of men, women and children in each house and the language practices of each individual.

Another reason for the delay on the interviews of the outlying communities has been bad weather, which makes paths more difficult and hazardous to walk, but we took the opportunity to start up the second part of the project, the recording, transcription and annotation of various kinds of dialogue and monologue, that is, the language documentation proper. So far we have spent about a week transcribing and glossing a story of about 7 minutes. This sounds like a long time to spend on 7 minutes, but it should be taking into consideration that at this stage, for every new lexical item, that is, for every verb, noun, adverb etc. (the word classes are still not firmly stablished, but I have a hypothesis that there are no adjectives in Tol, or rather, there are no words which modify nouns in the same way adjectives do in Indeeuropean languages; this function rather seems to be carried by either nominal or verbal predicates - so e.g. the words for 'bad' and 'many' are used without the presence of a modified noun and may take affixes which have otherwise been assumed only to occur with nouns. However, much more work needs to be done in order to form a proper analysis.

To return to the transcription process, as the language becomes increasingly familiar to me and the process of glossing and elicitation becomes more familiar to Federico, I expect the pace to pick up so that we might aim at transcribing various hours of speech, depending on the genre, since some genres are more text heavy than others - narrative tends to be heavy whereas small talk or childs play will more silence and repetitions.

Time will tell ...

23 interviews and counting

After these later interviews a new pattern is starting to take shape and some of the other information that I got earlier is making even more sense now. As a preliminary observation I would say that the population of speakers is not as homogenous as I thought earlier and there seems to exist a correlation between socio-economical factors and degree of language syncretism, i.e. the influence of Spanish - basically the way you speak Tol as well as Spanish tells a lot about how much power you have, both in terms of pure economics but indeed to a large extent prestige. These are preliminary observations which, if valid for the population at large, have implications for the future of this language only some of which may be possible to predict.
The grand total of interviews is going to be 30, the final 7 of which will be with speakers from outside of San Juan, the main community.

Starting up

It has been a little over two weeks since I got the Tol documentation project up and running out here in San Juan and, as so often happens during any kind of project in the field, a number of adjustments have been necessary, but the overall plan remains the same and things are running as scheduled, somewhat to my surprise.


The first thing that struck me upon arriving in the mountainous valley where the rivers and streams cut through the landscape and pine tress stand majestically covering every slope and hilltop is how peaceful and pretty this little pocket of Honduran forest really is. Also, I much appreciate the relatively low temperatures (22-27) offered by the highland climate even though lately it has also given us plenty or rain!

I am going to divide things up into themes as shown by the tabs in the top menu – this will hopefully give you access to just the kind of information you are interested in.

There is a theme called “Landing on The Moon” which talks about my experience of culture shock which has been quite forceful in spite of my ample experience living in these parts of the world and in Honduras in particular.

Another theme called “The Language Survey” tells about what discoveries I have made so far and which preliminary analyses seem to present themselves. It also talks about what difficulties I have run into and what unexpected situations have emerged.

A theme entitled “Recording” treats the issues I have come across when attempting to achieve the best possible audio and video quality while minimizing the intruding effects of the presence of recording equipment and not least my own presence.

This will be all for now, but more is to come...

Still underway

As the project begins on 1 Oct. and progresses this blog will come to life ...