Abstract: LFG has a number of features that make it an attractive and useful framework for grammatical description, and for translation. These include the modular design of the system, the literal representation of word order and constituency in c- structure, a typologically realistic approach to universals (avoiding dogmatic assertions which make the descriptive task more difficult), and a tradition of taking grammatical details seriously.
(This talk was presented as part of a panel entitled “Directions of LFG: Many Paths”. Each of the six panel members was asked to “explain how LFG has related to their past work and to what they are doing now in their careers.”)
Since leaving Stanford a lot of my work has been focused on training people to do field linguistics, so I have not actually been using the full LFG formalism in my daily work most of the time. But I have found that the conceptual structure of LFG provides a very good framework for grammatical description, and that is what I would like to talk about today.
A number of people have asked me at various times whether SIL still teaches Tagmemics to its field workers.1 The short answer is “no”; there is a fair bit of variety from one place to another, but none of the major training programs currently use Tagmemics as their basic model. For those of you too young to remember Tagmemics, I might summarize it by saying it was Kenneth Pike’s attempt to extend the methods of structural phonemic analysis to morphology and syntax. The phonologists among us may be eager to point out that structural phonemics was not a very satisfactory model of phonology either; but in its day it was considered a great triumph, the envy of the other social and behavioral sciences. And for all its limitations, Tagmemics was very successful in one important respect: using this framework, Pike was able to get people with fairly limited training in linguistics to study and describe languages that had never been studied before.
This kind of primary fieldwork on previously undescribed languages is a difficult thing to do, and I do not believe it has gotten nearly enough respect in American linguistics during the past 40 years or so; so I do not want to minimize in any way the contribution of Pike and his students and colleagues. But I would
1 “SIL” originally stood for “Summer Institute of Linguistics”, but now only the initials are used. For information about the organization, see www.sil.org.