Skip to main content
Language Resource Center

Events of 2006-2007

Sep 25
Monday, September 26, 2006
Cognitive Science Meets the Language Learner:
Language Patterns and their Transgression in Sentence Comprehension"
Noyes Lodge
Luca Onnis, Cognitive Science, Cornell
Reception starting at 3:00, talk from 3:30 to 5:00 with reception following

The emerging area of Cognitive Science attempts to bring together insights and methodologies from different disciplines to understand human cognition. In particular, understanding human learning and language learning has played a central role in current cutting-edge research. One of the most exciting aspects of cognitive science involves a cross-fertilization of ideas from different disciplines. In this talk we provide an example of such cross-fertilization, by showing how insights from Applied Linguistics can be fruitfully explored and deepened using the tools of Cognitive Science (e.g., experimental methods, computer analyses of large texts of actual language). We argue that this type of basic research aimed at understanding how people learn and process their first and/or second language may, in turn, have direct beneficial effects for learning and teaching practices involving second language acquisition.

To illustrate our case, we present a project that is underway in the Psychology Department at Cornell University. The working hypothesis is that fluent sentence comprehension can be achieved because proficient native speakers possess knowledge of units of meaning larger than the word (Sinclair, 1996). Statistical analyses of large databases of real language suggest that most words in a language occur in patterns that display clear preferences for which company they want to keep. For instance, in English, the verb 'provide' typically precedes positive words (e.g., provide work) whereas cause typically precedes negative items (e.g., cause trouble). We hypothesize that these statistical patterns form units of meaning that imbue lexical items, and their argument structures, with semantic valence tendencies. We further propose that native speakers may possess productive and generalized units of meaning like ‘cause + negative event’. Thus a reader may be more fluent at reading a sentence like 'The television news caused pessimism among its frequent viewers' than a sentence like 'The television news caused optimism among its frequent viewers' because the former sentence conforms to the natural valence tendency of the verb cause. This may be so even if the specific word pairs 'cause pessimism' and 'cause optimism' have a low co-occurrence frequency because the pattern is productive and generalizable to a variety of words. This work builds on the notion that words occur in context (Firth, 1957), and thus establish selectional restrictions that speakers capitalize on to boost fluency in language processing.

A related question in bilingual studies is how to account for proficiency differences between early and late bilinguals. Several studies have documented that even very proficient second-language speakers lag behind native speakers specifically in the degree of knowledge of language-specific selectional restrictions (e.g. take a picture in English but do/make a picture in Italian, fare una fotografia; cf. Howart, 1998; Onnis, 2001; Towell et al., 1996). Here we propose that part of the difference between early and late bilinguals’ fluency in sentence comprehension may be in the processing of extended units of meaning such as ‘cause + negative valence tendency’.

The broader scope of this work is thus to contribute to teaching practices that develop native-like fluency in second language learners. The research may have direct implications for instructional methods used in second language teacher development, as being aware of statistical tendencies toward differential semantic valences for certain cognates may greatly improve second-language fluency. The research will also be informative for the further development of automated translation systems.
(a fuller statement)
Some references on corpus-assisted langauge teaching and learning:
2004. G. Aston, S. Bernardini & D. Stewart (Eds.). Corpora and language learners. Amsterdam: Benjamin.
2002. G. Aston. "The learner as corpus designer". In B. Kettemann and G. Marko (eds) Teaching and learning by doing corpus analysis. Amsterdam: Rodopi. 9-25.
2002. G. Aston. "Getting one's teeth into a corpus". In M. Tan (ed) Corpus studies in language education. Bangkok: IELE Press. 131-144.

More Corpus and concordance information


There are many approaches and techniques for using a corpus/concordance approach in language teaching. For a start, try this short essay from Athelstan, which gives some examples to show why and how a teacher might want to use a concordancer.
There are several ways of approaching language investigation of this type. If you have a corpus (a large set of searchable language) you will need a concordance program to query it and analyze it. Wordsmith, unfortunately available only for Windows machines, is probably the best know full featured program. Another quite powerful program is Monoconc Pro, also only for Windows. It is a bit more user friendly and not quite as powerful. Here is a comparative review of these two programs. I have a copy of MonoConc Pro, which anyone can come and try at the LRC. I have a line on a Mac concordancer, but I need to investigate its setup (involves running Python).
But there are also corpora (the plural of corpus) available online, together with concordance interfaces. Some of these are quite powerful. The Leeds site can search corpora in 9 different languages (English, Chinese, French, German, Italian, Polish, POortuguese, Russian and Spanish) at a wide range of complexity, from one simple word to powerful "regular expressions." These are derived from Unix programming and are worth learning if you need to do complicated searches frequently. The Leeds site also has (complex) information about creating your own corpus from the web.
There is a French-only corpus that can be searched online, in various ways. There is also an instruction/information page about this tool and a couple of search input pages for this corpus. Another online concordance site from Hong Kong allows searching on texts in English, Chinese, French and Japanese and is quite easy to use. It also has some other straightforward tools and articles.
There is an impressive Spanish-only online concordance site done at BYU. This allows selection of any of eight different centuries of texts, including oral texts from the 20th century. You can access many of the original full texts. It is a powerful site, but also easy for beginners.
There appears to be a powerful German concordance site that gives broader collocations and statistics. It also offers a set of sophisticated synonym tools (I think).
Many of these corpora are "tagged" for grammatical function. You can then use grammar categories in your search. The Leeds corpus does this, as well as a German corpus available for free (but not searchable online as far as I can figure out).
I have a corpus of transcriptions of oral conversations in French from Betsy Kerr at Minnesota. She also sent me 11 CDs of the corresponding audio, though they are not automatically linked. In all, it's about 150,000 words. This is somewhat small as full corpora go, but very usable for many teaching purposes. You would need one of the concordance programs above to search it, though of course you can do simple word searching in a word processor. I can send/make copies of this for anyone interested. Betsy would just like to hear of interesting uses people make of the material. she has her own site of information and bibliography. There is another French corpus available from the University de Bretagne Sud.
The LRC will devote a workshop to trying out concordancers some time later this year.
Nov 1
Wednesday, November 1, 2006
Noyes Lodge
Reception starting at 3:30, talk/workshop from 4:00 to 6:00 with reception following

Teresa Pica, University of Pennsylvania
"Content Based Approaches in Language Education: Contributions, Concerns, and Innovations"

An opening lecture will first describe and compare content based approaches to language education, with attention to the four most widely used approaches, i.e., immersion, adjunct, theme-based, and specific purposes. It will then review research findings that reveal contributions and raise concerns about the sufficiency of these approaches in meeting the linguistic needs of language learners, especially for grammatical features that are difficult to notice on the basis of content and communication alone.

Using English as an example, an innovative approach will be described that draws learners' attention to these difficult features within the context of content and communication. This approach can transform a text taken from a content-based curriculum into a focused, goal-oriented task that guides students to notice relationships between difficult to learn grammatical features and the meanings and functions they encode. Several task models will be presented, together with classroom data that reveal ways in which the tasks help students to notice features such as noun articles, pronouns, connectors and verb morphology as they encode meaning in their content texts.

The lecture will be followed by a demonstration and workshop on the design and development of content-based, language focused tasks that draw students’ attention to features of language that are difficult for them to learn on the basis of content and communication alone. The demonstration will reveal how language teachers can do the following: (1) identify language features that their students are developmentally ready to learn, but have difficulty noticing on their own; (2) choose texts that contain these features; (3) modify the texts to make the features more noticeable; (4) incorporate the texts into tasks that require students to notice the features in order to reach a successful outcome. For the workshop, participants are encouraged to bring their own content texts for task design and development, and to allow for authentic application of the task demonstration to classroom practice.

Professor Pica has recommended an article for background reading: Teresa Pica, Hyun-Sook Kand, and Shannon Sauro. " Information Gap Tasks: Their Multiple Roles and Contributions to Interaction Research Methodology." In Studies in Second Language Acquisition, 28: 2, June 2006.
Nov 14
Tuesday, November 14, 2006
Noyes Lodge
Reception starting at 4, talk from 4:30 to 6:00 with reception following

Munther Younes, Department of Near Eastern Studies
"A New Multimedia Text for Beginning Arabic"
Munther Younes, the head of Cornell's Arabic program, will describe the pedagogical assumptions; the balance of culture, language practice and metagrammar; spiraling of vocabulary and other structures; the process of creating and producing dialogs and media; and student reaction to his material. This book is about to be published in its second edition, with accompanying audio CDs and video DVD.
Feb 13
Tuesday, February 13, 2007
"Cognitive grammar and narrative structure: Approaches and Tasks for Tense/Aspect"

Carl Blyth, University of Texas, Austin

Noyes Lodge
Reception starting at 3:30, talk from 4:00 to 5:30 with reception following
Carl has sent some suggested readings, two of which I have put online, as well as a group of online references related to his talk.

Abstract:
In this talk, I will demonstrate the application of Cognitive Grammar to the development of a coherent approach to the teaching of narrative past tenses. In general, cognitive grammar seeks to examine the relation of language structure to things outside language itself: general cognitive principles and mechanisms, including principles of human categorization and visual perception; pragmatic principles such as presupposition; and functional principles such as iconicity and economy. Central to cognitive grammar is the figure/ground distinction that originated in Gestalt psychology studies of visual perception. I will discuss the relationship between the figure/ground distinction and the perfective/imperfective aspectual distinction (e.g., preterit vs. imperfect). In the second part of the talk, I will derive and demonstrate pedagogical principles for teaching narrative past tenses based on SLA research and cognitive linguistic research about how the mind perceives visual information and construes that information as events in a narrative. Various techniques and activities taken from college French materials in use at the University of Texas at Austin and the University of Chicago will be demonstrated to make these ideas as concrete as possible. The approach to the teaching of aspect illustrated in these materials contains the following features:
  • Inductive learning
  • Video and visual mnemonics to help learners construe narrative events
  • Transparent cinematic metalanguage to describe aspectual distinctions
  • Careful pedagogical sequence: from input activities to guided output activities
Dec 05
Tuesday, December 5 9am - 2pm
Lunch included
LRC Winter workshop
The workshop included demonstrations and hands-on practice with a range of new tools developed by the LRC, including means to download videos from various sources, upload them to wrkbooks and display them with a video iPod.
Mar 13
Tuesday, March 13, 2007
"Language and the Media Courses"
Ute Maschke and Elvira Sanchez-Blake

Noyes Lodge
Reception starting at 3:30, talk from 4:00 to 5:30 with reception following
Courses here and at other schools have been devised recently exploring media sources in the target language. The two presenters have several years of experience with such courses. Though they have conducted their courses in rather different ways, they share a strong emphasis on both technology and collaborative student presentations. Ute and Elvira will discuss their courses, demonstrate some of their materials and show some of the student presentations. Ute's course was awarded funding from the Faculty Innovation in Teaching Grant program.
Mar 23
Friday, March 23, 2007
"Outreach Workshop: Best Practices in Language Pedagogy"
Dick Feldman, Ute Maschke and Silvia Amigo-Silvestre

Noyes Lodge
9:30 - 2:45
This all-day workshop will include 4 talks: "The role of explicit instruction on language form in task-based teaching" and "Download, edit and play video from web sources like YouTube and Google Video" by Dick Feldman; "Teaching and Learning - Assessment and Self-Evaluation" by Ute Maschke; and "Principled and creative tasks with songs in class" by Silvia Amigo-Silvestre.
Dick's presentation, the Quicktime editing instructions the instructions for downloading from YouTube prepared by Nick and a related article are available online.
Mar 30 Mar 31 Apr 01
Friday, March 30, Saturday, March 31 and Sunday April 1
Noyes Lodge and Purcell Union
This year the annual meeting of the Northeast Association for Language Learning Technology will be held in Ithaca on the Cornell Campus. This conference is generally attended by 50 or so people from around the Northeast. For an idea of the topics, see their recent newsletter. The conference includes a Friday afternoon workshop sponsored by the Cornell LRC at Noyes, then a plenary and 3 concurrent sessions held at Purcell Union on North Campus on Saturday and Sunday. The topic for 2007 is "The Language Learning Space: Real and Virtual Uses." More information on related events and registration will be coming soon.
Apr 18
Wednesday, April 18, 2007
Reception at 3:30, talk from 4 to 5:30, with reception following

"The effect of trait emotional intelligence and sociobiographical variables on communicative anxiety among adult L2 learners and users"
Jean-Marc Dewaele
University of London, Birbeck
Abstract: This study considered the effects of trait emotional intelligence (Petrides & Furnham, 2001) and sociobiographical variables (age, gender, education level, age of onset of acquisition, context of acquisition, frequency of use, socialization, network of interlocutors, self-perceived proficiency) on communicative anxiety (CA) in the first, second, third, and fourth languages of multilingual individuals, in five different situations (speaking with friends, colleagues, strangers, on the phone, and in public). Data were collected via web-based questionnaires completed by 464 (123 males) adult polyglots, who were divided into three groups based on their trait emotional intelligence scores (low, average, high). Non-parametric statistical analyses revealed a consistent pattern of results across languages and situations. Higher levels of trait emotional intelligence corresponded with significantly lower CA scores. Participants who started learning the L2 and L3 at a younger age also tended to suffer less from CA in certain situations. Those who had learned a language through formal instruction only suffered more from CA than those who had used the language outside the classroom. A higher frequency of use, a stronger socialization in a language, a larger network of interlocutors and a higher level of self-perceived proficiency in a language were also linked to significantly lower levels of CA. Findings are discussed with reference to earlier research on polyglot samples and ramifications for language teaching.

We have an audio file of this talk and the accompanying Powerpoint file. Please contact Dick Feldman (rf10) if you would like a copy.
May 8
Tuesday, May 8, 2007
LRC Spring Workshop
9am to 2pm with lunch provided
Noyes Lodge
Similar to last May, this workshop will be a combination of discussion of impending changes at the LRC, demonstration of new features in LRC tools, and a chance to review how the Account Manager makes management of workbooks, media uploads and course pages easy and organized.
Many of you attended the NEALLT conference several weeks ago. I'd like to have some discussion of specific ideas or tools, or possibly more general themes you heard about at the conference. This will also serve as information to those who did not attend.
Related to this topic, I am on a CIT committee reviewing a number of course management systems as possible alternatives to Blackboard. I will report on that.
The environment has somewhat changed regarding the scope of fair use of commercial video on websites. I'll discuss this, and the related issue of other possible video sources, such as dishNetwork. There is a great deal of material available on dishNetwork, in most of the languages taught at Cornell. The LRC could make programs available, time-shifted, for teachers to use in materials development. This would not affect our SCOLA subscription, but would augment it, with a greater variety of program types. Take a look at the international and educational (meaning special pricing for educational institutions) packages available.
Some other questions I would like discussion/input on: Is there demand for a high-level digital audio/video editing station for student use? Are there other tools that would be useful for teachers and students?
Update on LRC changes: still working on in-house video streaming; planning furniture changes in the circulation area, mostly reflecting the end of any volume of cassette usage.
Gunhild Lischke has made an excellent suggestion for our next workshop in December: panels of Cornell teachers talking about areas of shared local concern and practice. These would not be prepared presentations but rather an opportunity for discussion about what teachers do in different programs. Here are some suggested topics:
  • Credit from outside CU study
  • Heritage stduents
  • New courses - how to develop
  • Uses of classroom equipment
  • Scheduling with several teachers for each section
  • Student-produced video

We will have some exciting news about tools for podcasting. We'll demonstrate that, give attendees a chance to try it out, and take suggestions on how it should be developed further.