Department News and Events

The Ohio State University



General Information
The Ohio State University Department of Linguistics is pleased to offer weeklong workshops led by OSU faculty and researchers. The Mini-Institute will take place July 14-18, 2008, immediately following the LSA 2008 Summer Meeting. Each workshop will meet on the OSU campus for 1½ hrs each day, Monday-Friday. Participants may attend both the Meeting and the Workshop, or just one event. Note that registration for the Mini-Institute and the LSA Summer Meeting are independent.

The Mini-Institute course schedule is set up to allow participants to attend more than one workshop, if they so choose.

Workshop times and instructors:

9:00-10:30: Quantitative methods for linguistics - Mary Beckman (Workshop Full)

11:00-12:30: Corpora and Corpus Studies - Chris Brew & Mike White (Workshop Full)

1:00-2:30: ToBI labeling - Julie McGory

3:00-4:30: Eye tracking methodology - Kiwako Ito & Shari Speer (Workshop Full)

5:00-6:30: Methods of Analysis in Sociophonetic Research - Robert Fox, Ewa Jacewicz and Joseph Salmons (Workshop Full)

For more information on the instructors, visit the Instructor's Bio Page.

Fees
The registration fee is $200.00 (US$) for participants enrolled in up to two workshops, and $300.00 (US$) for those enrolled in more than two workshops.
1-2 workshops - $200 (US)
3 or more workshops - $300 (US)

Registration deadline: Wednesday, April 30, 2008 5 p.m. EST

Registration Information
Registration for this event is no longer available.

Accommodation
Meals and accommodation in university dorms and are provided free of charge for student participants (undergrad or grad students or non-students entering graduate school in fall 2008). Dorm rooms are double-occupancy and air-conditioned. Participants choosing to live off-campus will be responsible for locating and paying for housing themselves.
Information will be sent to registered participants.

Travel
Registered student participants can request travel support according to the schedule below:

Mode -- Distance -- Amount
Car/Bus - $100 max
Air   - Ohio (except Columbus) - $125 max

Information will be sent to registered participants.

---

The Ohio State University 2008 Linguistics Mini-Institute

Workshop Descriptions

Quantitative methods for linguistics - Instructor: Mary Beckman - Section Full
A familiarity with quantitative methods for analyzing language data and a basic facility in interpreting numerical models of language phenomena have become increasingly important to linguists today. Experiments and speech synthesis models have long been central in the development of phonetic theories, and they are becoming more and more central in the development of theories of phonological structure and processing and their relationship to sound change. More recently, arguments from experiments and models have been central to the evaluation of competing models of the phonology-syntax interface in languages with rich systems of inflectional morphology for nouns. Also, as syntacticians have come more and more to realize the "limits of syntax", experiments, corpus studies, and computational parsing models have come to play a central role in syntax and semantics as well. This course is a very basic introduction to some aspects of quantitative methods in linguistics. Two small case studies will be covered, with copious notes and references to other resources that students can use to begin to build the necessary skills to understand numeric arguments.

ToBI Labeling - Instructor: Julie McGory
ToBI (Tones and Breaks Indices) provides a framework for intonational transcription that is used widely in both speech technology and linguistics research. Intonation interfaces with multiple areas of linguistics (including sociolinguistics, syntax, pragmatics, psycholinguistics, and many more) providing necessary information to the interpretation of language. Consequently, understanding the role of intonation in language is fundamental to technological advancements in speech recognition and synthesis, and to understanding linguistic phenomena. For this reason, there is a growing need in the linguistics community to acquire ToBI transcription skills. While one can read the literature associated with ToBI and study the guides that are readily available, it is difficult to acquire a working knowledge of English intonation and its transcription without actively participating in an applied workshop such as this. The questions that one asks, the practice one receives, and the mistakes that are made are all necessary to grasping the ToBI transcriptional system. The goal of this workshop is to provide participants with a working knowledge of the transcriptional notation needed to describe the tonal movements in read and spontaneous American English speech. In this week-long course, we will review the key components of the ToBI transcription system and discuss its theoretical basis while at the same time labeling progressively more complex English utterances.

Corpus-based computational linguistics - Instructors: Chris Brew & Mike White
Standardized tasks and shared corpora play a crucial role in modern computational linguistics. Because the tasks are standardized, the cost of replicating and extending previous studies is a significantly reduced. Because the corpora are shared, progress depends not just on the abilities of a few groups with privileged access to data but on the talents of the whole of the relevant research communities. All of this is clearly beneficial. Wider access to data magnifies our capabilities and promotes a level of care and rigor which advances the field. To illustrate this, we will present a longitudinal study of the various ways in which corpora derived from and otherwise related to the original Brown Corpus have been used to ask and answer scientific questions.
While many computational linguists learn to use corpora and appreciate their strengths, relatively few learn about the design principles which make them more or less useful in the longer term. It is easier and more common to make uncritical use of "generallly accepted corpora" than to carefully and critically think through the reasons why the corpora are the way they are. The second aim of the course is to provide some of the tools that will allow students to understand and adopt the designer's perspective, and thereby to understand which questions can reasonably be asked and answered about these corpora. We hope to provide a basis for intelligent decisions about adopting, adapting and creating corpora that are suitable for particular scientific questions.

Eye tracking - Instructors: Kiwato Ito and Shari Speer - Section Full
Eye tracking is one of today's most popular techniques in psycholinguistics. Thanks to constant mechanical advancements, various eye tracking systems have become widely available and more easily learnable in the past few years. This non-invasive online measurement attracts those who investigate how people process sentences during reading text, as well as those who study comprehension and production of speech.
Eye tracking is replacing many traditional psycholinguistic methods such as self-paced reading, word or phoneme-monitoring, lexical decision and sentence verification tasks for a variety of reasons. First, an immediate temporal link between language processing and eye movements has been established, and thus the technique allows online observation of language processing. Second, the participants are generally unaware of when and where they are sending their own eyes, therefore the data are less contaminated by experiment-internal strategies. Third, as the eye fixation duration, fixation proportion, saccade latency and frequency all convey information about the timing and the degree of processing ease and difficulty, unnatural extra tasks such as pressing a button for reading the next word and monitoring words or phonemes while listening to speech are not necessary. This makes the technique ideal also for studying children and second language learners, whose data tend to reflect artifact of experimental tasks. Through observation of eye movements during a normal language activity, investigators can more readily draw generalizations about natural language processing.

Methods of analysis in sociophonetic research - Instructors: Robert Fox, Ewa Jacewicz and Joseph Salmons - Section Full
Sociophonetics has emerged as one of the most active areas of linguistic research in recent years, best known for the quantitative analysis of variation in regional dialects, speech styles, or speaker groups. The field applies methods used in phonetic and speech communication research to the exploration of social factors in speech and language acquisition with a particular interest in explaining the transmission of linguistic sound change. This workshop explores fundamental ways that speech science and historical phonology connect with sociophonetics, while introducing innovative methods and experimental techniques being used in current work on regional dialect variation and change. Topics will include measurement and statistical assessment of acoustic properties of vowels, their cross-generational transmission, speaker normalization and estimation of a working vowel space, comparison of data from spontaneous speech and more structured phonetic elicitation, and the organization and management of an extensive database. In addition to attention to social correlates of language variation, we will explore how this work can inform our understanding of classic problems of sound change, including directionality of change and perseverance of vowel shifts over many generations. While the focus will remain on principles, we will draw data from a number of different American dialect areas.