legacy materials

[William J Poser]

Phil Cash Cash writes:

>I should add that the recent trends in the linguistics field are
>focused almost exclusively on the creation of legacy materials
>and less so on current archived materials despite their relatively
>equal status.

Two comments:

First, there is a very good reason for this: when we're dealing with dying
languages, it is important to gather data now, while they are still
alive. Analysis of legacy materials can be done in the future, when
there are no more native speakers; gathering of new material cannot.
In a world with infinite resources, we could do both, but in the real
world, with very limited resources, time spent studying legacy materials
is time not spent gathering new material.

Indeed, for this reason some of us have made conscious decisions about
where to place our priorities. In my case, while I am quite interested
in historical linguistics, some time ago I made the decision not to
spend very much of my time on it because it is something that someone
else can do in the future.

Second, generally speaking legacy materials and new material do not
have "relatively equal status". Of course, the relationship depends
on exactly what legacy materials are available and what the current
state of the language is. If the language is still in sufficiently
good condition to yield copious new data, it is very likely that
legacy materials will be inferior to new material for reasons
including the following:

(a) legacy material is often poorly transcribed. Relatively recent
    material recorded by professional linguists is likely to be accurate,
    but material recorded by non-professionals, such as missionaries,
    fur traders, and, often, anthropologists, is often poor. Earlier
    materials recorded by professional linguists, insofar as there
    were such things, is often not very good, or, even if the linguist
    heard well, may be very difficult to interpret due to the lack of
    a standard notation at the time. (Harrington's work is a major
    exception - he had an unusually good ear and his transcription
    was very accurate, though it does pose difficulties of interpretation
    since he was known to do such things as switch notation in the
    middle of a page of notes.)

(b) legacy material is often restricted in genre. For example, in
    one major anthropological tradition, the principal activity
    was to collect texts, where the texts usually consist primarily
    if not exclusively of legends and/or oral history. Such texts
    are often of great cultural and historical interest, and they
    do yield insight into the language, but they are also quite
    defective as a source of information. For one thing, such texts
    are often in a form of the language different from the ordinary
    spoken variety. In many cultures there are special conventions
    for telling such stories and the language is often archaic.
    One misses the language of ordinary conversation, of speeches,
    of prayer, etc. One may even miss common grammatical forms.
    For example, narratives may contain few or no 1st and 2nd person
    forms since everything is told in the third person.

(c) linguists of earlier times tended to focus on lexicon and morphology
    but to have very little to say about syntax, semantics, and
    discourse. 

On the other hand, if the language is in poor shape, the material
that can be obtained from the last few speakers may itself be
limited. The last "fluent" speakers often have a narrower range
of genres than their parents or grandparents and may also have
a narrower range of vocabulary due to the decreasing range of
circumstances in which they use the language. Their language
may even be "degenerate" (no moral judgment intended here) in
exhibiting such effects of language death as simplification of
morphology and loss of phonological distinctions.

In sum, where the language is still in sufficiently good shape
as to provide good data, it generally  makes sense to devote
limited resources to gathering new material at the expense of
analyzing legacy material.

Bill