This course discusses the common confusion between different 
approaches to linguistics and language study, and provides a 
general theoretical framework under which empirically-based 
investigation of speech can be carried out. This course will 
first step back from the modern practices of linguistics in 
order to examine the choices available for language study, as 
usefully categorized a century ago by Saussure.  Then, the 
course will consider empirical evidence from real speech, 
primarily from survey research and corpus linguistics but also 
from other empirical studies of language in use, to answer the 
question "What model of human language does this evidence lead 
us to build?" In particular, such study will introduce students 
to language variation and will consider how the fact of 
variation should condition how one thinks about language. 
Language as people use it creates expectations among 
professional linguists and the public that have strong 
implications for social and educational policy. All students 
will be expected to produce at least 20 pages of writing about 
the issues of the course, at a level suitable for professional 
use at conferences and in publications. Students will be 
expected to become familiar with the relevant literature, 
including the ability to cope with statistical arguments. 
Students will be required to write at a much higher level of 
professional competence than undergraduates, even though 
undergraduates also have to write intensively. Students are 
also required to review and become conversant with the 
published analytical literature in the field, while 
undergraduates are not. Finally, students are required to 
understand and evaluate statistical arguments that are common 
in this field, while undergraduates are not.

Particular sub-topics will vary from semester to semester, 
especially to keep pace with emerging new research in empirical 
linguistics.

1.  Introduction.  The course will begin with the perennial 
interest of the public in language variation and document some 
problematic cases of variation, such as the Ebonics debate and 
language standards in education.

2. Saussure.  Discussion will turn to Saussure's decision to 
prefer the "linguistics of linguistic structure" as opposed 
to "the linguistics of speech." The course will revisit 
Saussure's description of the "linguistics of speech" and its 
separation from other relevant fields (prescription, philology, 
psychology/anthropology). This section will also differentiate 
the principal modern modes of linguistics (structural, 
generative, Firthian), including discussion of dialectology and
sociolinguistics.

3.  Evidence from Linguistic Survey Research. This section will 
discuss evidence from linguistic survey research, for which UGA 
is a national center, including discussion of categorization, 
tabulation, and mapping of findings.

4.  Social Variability. This section will discuss statistical 
processing of variation data, including techniques that 
integrate multiple regional and social association of 
individual linguistic features.  

5.  Evidence from Corpus Linguistics.  This section will 
consider basic findings of corpus linguistics, primarily as 
developed in the work of Biber, Halliday, Sinclair, and 
Stubbs.  Particular topics include text types and collocation.

6.  Speech, Power Laws, and Models of Language. This section 
will address the question of what is "linguistic" about the 
distributions of language variants and will argue that speech 
conforms to the characteristics of complexity science as 
developed in the physical and natural sciences. Non-linear 
distributions and scaling will be treated as primary 
characteristics of speech distributions.

7.  Linguistic Perception. This section will address the 
findings of perceptual language variationists, such as Preston 
and Tamasi.  Further, the model of language production will be 
extended to include the creation of "observational artifacts" 
that correspond to "language systems" as traditionally 
associated with "speech communities," and to address cognitive 
models for language.

8.  Applications.  This section will consider applications of 
the complex systems model for speech that emerges from study of 
language variation, including the relationship between language 
variation and language change, and speech and public policy.

Students in this course are expected to be familiar with and
adhere to the University of Georgia policy on academic honesty,
according to which all violations of academic honesty will be
handled.  Stuents may participate in graded group projects at 
the instructor's discretion.