Ruth Singer recaps some of the interesting points of the last week’s Linguistics in the Pub, an informal gathering of linguists and language activists that is held monthly in Melbourne
The announcement for this month’s Linguistics in the Pub outlined the topic as follows:
The aim of language documentation, broadly speaking is to document linguistic diversity. At one level the diversity refers to the range of languages and dialects that are used. But zooming in a bit closer diversity can be understood to refer to the variation in how language is used across different speakers and contexts, i.e. social variation. Despite the close link between linguistic diversity and social variation, variation is often viewed mainly as a problem in initial stages of documenting and describing a language. It is more challenging to describe a system of phonology, grammar or morphology when it varies widely, than to describe a system with little variation. For this reason, it is often only after documenting one variety that linguists usually try to document broader socialvariation and patterns of language use. In this session, we will look at some good examples of documentation of linguistic variation and discuss how we might include some aspects of social variation in language documentation projects right from the start.
The topic was initially planned for February LIP but for we ended up discussing other things in February so we revived the topic for April to coincide with a visit Felicity Meakins had planned to Melbourne. There was a great turnout thanks partly to the participation of 4 students from Monash University working in sociolinguistics. Felicity’s involvement also no doubt attracted a good crowd due to her interesting research that combines research on social variation in the context of language shift and the development of a new mixed language, with language documentation.
The discussion was structured around issues related to how we can include awareness of sociolinguistic variation in language documentation identified. As the starting point for the discussion we used two lists of issues discussed by Naomi Nagy in Nagy (2009) – The challenges of less commonly studied languages: writing a sociogrammar of Faetar (see February announcement for details of the three background readings). Nagy first discusses the problem of efficiently and effectively codifying a language without misrepresenting the facts of language change and variation. Most of us at the discussion who are carrying out language documentation prefer to use a standard orthography and spelling for our transcriptions, despite variation in actual pronunciation, whether this represents social variation alone or connected speech phenomena. However, as Nagy points out this means we are obscuring much of the phonetic variation in our data. A database with consist spelling means we can easily search the database for lexemes, grammatical constructions etc. However it means we need to retranscribe sections of our database if we wish to study sociophonetic variation, although some brave souls include both a standardised transcription and a more phonetic transcription in their annotations. The idea that Nagy has pursued in her documentation of Faetar is that there is no need for a standard consistent format for transcriptions. This idea links to the May 2011 LIP session led by Felix Ameka which discussed whether language documentation requires language standardisation. He suggested that there is an advantage to a variable writing system which allows people from different dialect groups to be able to write their language as they speak it, without the need to memorise unnatural spellings that do not reflect the way they pronounce words. However it may take a lot of variation to convince those of us who carefully tend our interconnected databases that it is a good idea to give up on consistent spelling. This will make it more difficult to link a dictionary with a time-aligned text database and eventually it should be easier to link these to a grammar too.
The other main topic we discussed was how different the task of writing a grammar is from most sociolinguistic work. They differ in terms of the number of speakers recorded, the number of linguistic phenomena studied and also the type of data recorded. In sociolinguistics we aim to record a wide range of speakers, even if we are restricting ourselves to one particular subgroup of a speaker community. In addition sociolinguistic research tends to focus on a single linguistic phenomenon and see how it varies across speakers, whereas a descriptive grammar aims to cover as many topics in as many areas of structural lingustics as possible (e.g. phonetics, phonology, morphology, semantics, syntax, discourse etc.). The third main difference is that sociolinguistics tends to prefer as natural data as possible whereas grammar writing needs some elicitation to supplement natural data or monologic narratives. Another point raised by the group is that field linguists working on grammatical description, or even transcription of texts tend to develop good working relationships with a few speakers who are very patient and willing to take part in the fairly dull and time-consuming business of helping a non-native speaker linguist understand a language from scratch. This means that the bulk of examples in a grammar may well come from a small number of speakers, and there are grammars that are based on consulting a single speaker (not that these are necessarily ideal). Generally even if linguists consult a wide range of speakers, they limit their grammars to a variety spoken in a set location, such as a village-based dialect.
So the group concluded that there is a limit to how much social variation can be included in grammars, and there may be costs to including much variation, such as that less topics will be covered in total. However, there are some things that we can do when documenting a language, to make sure that important sociolinguistic information is preserved along with out other materials in the archive: create a good speaker database. It has become standard to include speaker names in our transcriptions (for example in elan tier names), and this is now sometimes done also in grammars (a code may be included in example references). However it is important to include a biography of all the speakers recorded, making note of everything we know about the speakers that might be relevant such as their age, ethnic identification, where they live, how they are related to the other speakers recorded, if they are 1st or 2nd language speakers of the language recorded, etc. Another thing which Felicity mentioned was the importance of writing a sociohistorical sketch of the language and the speech community, as in many cases there are few little historical records of communities which speak endangered languages. Grammars often include this in their introduction, but not always.
The next Linguistics in the Pub will be on Tuesday the 29th May and will look at the ‘technique’ of participant observation – what is it, do linguists do it and should we start talking about it like anthropologists do? Information will be made available on the RNLD email list, events page and facebook page.