“A multilingual approach to analysing test results and the role of languages spoken in a bi-/multilingual community” by Gessica De Angelis

Leave a reply

This post was contributed by Agnès Marchessou, a PhD student in Birkbeck’s Department of Applied Linguistics and Communication.

linguistics_50_finalDr Gessica De Angelis’ seminar was the second in a series of talks being hosted by Birkbeck’s Department of Applied Linguistics and Communication, as part of its 50th Anniversary celebration. It gave us the opportunity to hear from one of the great alumni contributors to the field. De Angelis obtained her PhD at Birkbeck in 2002 with Professor Jean-Marc Dewaele as supervisor. She worked at the University of Toronto, the University of Bolzano and is currently Assistant Professor at Trinity College Dublin.

One of the most inspiring aspects of De Angelis’ research is the fact that it has sociopolitical implications, well beyond the field of Applied Linguistics. By presenting a critical overview of standardised tests, De Angelis exposes the language barriers to an equal education, with ramifications at policy-making level.

The research presented takes place in the province of South Tyrol (Northern Italy), a multilingual region where three languages coexist within the community: German (the largest population), Italian and Ladin (the Dolomites’ dialect). De Angelis took us through the traumatic and complex history of the region, torn apart by neighbouring nations. Of particular interest was the signing in 1972 of Article 19, the statute which gave the right to education in any of the three local languages, and introduced three separate monolingual school boards. By stipulating that teachers had to be of the same mother-tongue as their students, it actually made bilingual education illegal.

De Angelis presented two pieces of research in this complex sociolinguistic context.

The first study (De Angelis 2012; De Angelis and Jessner, 2012) looked at the relation between school performance and the language spoken in the community (population distribution), with speakers of Italian L1, German L2 and English L3. Written national evaluations undertaken by 14 year old students were used as data. The results showed a typical association between L2 exposure and proficiency, but most importantly, the research demonstrated that it was the absence rather than the presence of Italian L1 in the community that made the difference, by increasing proficiency in L2 German. In other words, a context with a majority of L1 German speakers gave the L1 Italian group more opportunities to communicate with them and therefore improved their L2 proficiency in the language of ‘the other’. This in turn had an effect on L3 competence, given that high proficiency in German L2 was found to have a positive effect on English L3.

The second study (De Angelis 2014) focused on the language proficiency of immigrant children attending Italian schools in Italian and German-speaking areas. National school test results on seven year olds were compared at a national and local level. Unexpectedly, in the case of South Tyrol, both 1st and 2nd generation immigrants underperformed to a similar extent, when it would typically be expected that 2nd generation immigrants would outperform 1st generation ones (as shown in national results). When reassessing the data in view of the language spoken in the community, 2nd generation immigrants actually also outperformed 1st generation ones in German speaking areas (where opportunities to practice Italian within the community are limited).

Consequently a multilingual approach can prove beneficial for the accurate reading of test results within specific local linguistic contexts. The misinterpretation of such results can have heavy financial repercussions for the use of (often limited) resources in education. Standardised tests also have important social implications (McNamara & Roever, 2006), therefore raising concerns about their interpretation may assist in addressing inequalities (in education for instance). I have found such insights into the language conflicts of South Tyrol most valuable, as my current research is taking me to another European border region, Alsace.

De Angelis, G. (2014) A multilingual approach to analysing standardized test results: immigrant primary school children and the role of languages spoken in a bi-/multilingual community. Intercultural Education, 25 (1), 14-28.

De Angelis, G. (2012) The effect of population distribution on L1 and L2 acquisition: evidence from the multilingual region of South Tyrol. International Journal of Multilingualism, 9 (4), 407-422.

De Angelis, G. and Jessner, U. (2012) Writing across languages in a bilingual context: A dynamic systems theory approach. In R. M. Manchòn (Ed.) L2 Writing Development: multiple perspectives. Trends in Applied Linguistics Series. Mouton de Gruyter. 47-68.

McNamara, T. F. and Roever, C. (2006). Language testing: The social dimension. Malden, MA: Blackwell.

See the full recording of the talk here.

The importance of Web Science

Leave a reply

Richard has a BSc in Physics from University of Leicester and an
MSc in Advanced Richard Brownlow copyInformation Systems from Birkbeck. He has over 20 years’ experience in industry as a Software Engineer and Software Project Manager and is currently studying for a PhD at the London Knowledge Lab where he is a member of the Weaving Communities of Practice Project. His research is in the design of tools to help domain experts integrate heterogeneous data sets.This post was contributed by PhD student Richard Brownlow. 

 

Annually at Birkbeck, the Department of Computer Science and Information Systems celebrates the work of its founder, the late Dr Andrew Booth, who was a pioneer in computer hardware and machine translation. Hosting this year’s Andrew Booth Memorial Lecture was the London Knowledge Lab, a unique interdisciplinary collaboration between two of the UK’s most prominent centres of research – Birkbeck and the UCL Institute of Education.

This year, we were honoured to have Professor Dame Wendy Hall present. She has played a foundational role in the development of the Web, the Semantic Web and Web Science, with her current research focussed in applications of the Semantic Web and in exploring the interface between the life and physical sciences. Along with being the first person outside of North America to be elected to the post of President of the Association for Computing Machinery (ACM), she has also been hugely influential and inspirational in promoting women’s careers in computer science.

Along with Professor Hall’s lecture, a broad range of London Knowledge Lab research was on show in the Department, for staff, students, alumni and guests from other institutions and across the industry. Opportunities for future collaborations and research were discussed. Some of the research demos included projects relating to Learning Technologies, such as LIBE which supports literacies through lifelong learning with inquiry based education. Other research demos were in the areas of ontology querying and mobile location analytics. I was also given the opportunity to demonstrate some of my own research interests including the knowledge base developed for the Weaving Communities of Practice project.

The importance of Web Science

The magnificent Keynes Library in Gordon Square was the setting as Professor Hall kindly delivered her lecture, captivating the audience with her insight on what the discipline of Web Science means in the context of the history of the World-Wide-Web. This was especially interesting given the foundational role she played in the development of the Web, including her collaborations with other giants of the sector such as Sir Tim Berners-Lee.

internet

Discussing the role of the Web in knowledge creation and sharing and the need to understand it in terms of both its technical and its social aspects, she also spoke on how this multidisciplinary field has come to be known as Web Science and the establishment of the Web Science Trust (WST) in 2006. She went on to describe how Web Science encompasses the theory and practice of Social Machines and how such machines are quite different from Turing Machines, which lie at the heart of every computer.

Professor Hall described the establishment of the Web Science Trust Network of Laboratories (WSTNet), an initiative furthering academic excellence in the field. There are currently fifteen such labs, including two in the UK. She then went on to describe a new exciting initiative called the Web Observatory, through which global partnerships are established to share data sets (both open and closed) along with associated Metadata and Analytics tools. Through these initiatives, Professor Hall described how Web Science aims to understand the origins, current state and possible futures of the Web, and to further the development of new research methodologies.

It is just over 10 years since Professor Hall delivered one of the inaugural talks at the London Knowledge Lab. In her vote of thanks, Professor Alex Poulovassilis – one of the two Co-Directors of the London Knowledge Lab – drew links to that inaugural lecture, firstly in the role of the Web in knowledge acquisition, sharing and dissemination, and secondly in the need to keep historical “memories” of the Web in order to enable the longitudinal analyses required for understanding its evolution and future.

Continue reading →

“Tasks vs. Conditions: Influences on Second Language Performance” by Professor Peter Skehan

Leave a reply

This post was contributed by Chieri Noda, a PhD student in Birkbeck’s Department of Applied Linguistics and Communications.

linguistics_50_finalWe had the pleasure of hosting a lecture by Peter Skehan, Professorial Research Fellow at St. Mary’s University, Twickenham on 27 October. Being the first Applied Linguistics student to complete a PhD at Birkbeck, Peter was the perfect speaker to kick off the series of talks marking the department’s 50th anniversary. The lecture hall was packed with attentive students and academics from not only Birkbeck but neighbouring institutions – ‘the beating heart of Applied Linguistics’ as the emcee put it.

After reminiscing about his days at Birkbeck, Peter started by giving an overview of the dimensions commonly used for measuring second language task-based performance (complexity, accuracy, fluency and lexis) and by highlighting the differences between the two contrasting approaches in cognitive research in this area, i.e. Cognition Hypothesis (Robinson, 2011) and his own Limited Attentional Capacity framework (Skehan, 2014), often referred to as the Trade-off Hypothesis. After pointing out some of the problems with the construct of task complexity on which the Cognition Hypothesis is built, he went on to establish his view that task conditions could better illuminate second language performance.

He took us through research on the effects of on-line planning, repetition and post-task planning he had done with his PhD students and co-researchers in Hong Kong. The studies can be found in a recent book he edited, Processing Perspectives on Task Performance. Particularly interesting was Zhan Wang’s 2014 study in which conditions of on-line planning (i.e. planning while speaking) were carefully controlled using slowed Mr Bean videos. The results suggested that on-line planning itself did not improve performance, but a combination of pre-task and on-line planning boosted accuracy and complexity. While at first glance this might seem to support the Cognition Hypothesis, Peter proposed that with the help of Willem Levelt’s (1989) psycholinguistic model of first language speech production, these effects could be explained to be due to two separate psycholinguistic influences, one of which raised complexity and one of which raised accuracy.

One thing that the audience had the privilege of hearing, which readers of Peter’s recent book would be missing out on, was the amusing but nevertheless insightful account of how his PhD students had been stimulated by past research and had developed their idea into well-designed research. Perhaps it was the accumulation of such experience that prompted him to share a piece of practical advice for MA and BA students in the audience. “In some ways,” suggested Peter, “the best thing to do is to find a study you half like. One that you like, but you’ve got reservations about.” He explained that students could redesign the study within the framework of the original study to correct problematic elements. He admitted this might be edging away from doing something brand new, but pushed students to master research techniques.

Peter’s talk was truly fitting as the first of our anniversary talks, not simply because it presented the findings of his extensive research over the years, but for showing us how he engaged with competing views and with his own data to drive his research forward. My doctoral research on Japanese researchers talking in English takes a conversation analytic perspective and so in terms of methodology very different from Peter’s, but his talk left me with a lot to reflect on.

  • Levelt, W. J. M. (1989). Speaking: from intention to articulation. MIT.
  • Robinson, P. (Ed.). (2011). Second language task complexity: researching the cognition hypothesis of language learning and performance. John Benjamins.
  • Skehan, P. (Ed.). (2014). Processing Perspectives on Task Performance. John Benjamins.
  • Wang, Z. (2014). On-line time pressure manipulations: L2 speaking performance under five types of planning and repetition conditions. In P. Skehan (Ed.), Processing Perspectives on Task Performance (pp. 27-62). John Benjamins.

A video of the lecture is available here.

Academia’s gender inequality problem

Leave a reply

This post was contributed by Professor Helen Lawton-Smith of Birkbeck’s Department of Management. Professor Lawton-Smith is organising Improving gender equality in work – what can we learn from London’s business and policy organisations? on Wednesday 18 March, 2pm-5pm.

laboratoryWomen are under-represented in senior positions in science, engineering, maths and medicine disciplines at UK universities. Initiatives including Athena SWAN and the Aurora Women’s leadership programme have been set up to address this problem, yet such initiatives by themselves are not enough to tackle the problem of the current gender bias. What is needed is institutional embedding, so that gender and other diversity issues are integrated into an equality framework of decision-making processes and structures within organisations, which cannot be side-stepped by those in positions of power.

The four-year Transforming Institutions by Gendering contents and Gaining Equality in Research (TRIGGER) project at Birkbeck is championing the role of female academics in scientific subjects as part of a five-country European project. This initiative is testing a blueprint designed to raise the status of women in scientific and technological organisations such as universities. The nine action areas are designed to identify barriers to equality in the workplace, including the impact of research. The project builds on Birkbeck’s existing commitment to promoting female academics. Results and reactions have been very interesting.

Equality issues have been tackled in a variety of ways by companies and by policy making bodies such as local authorities and government agencies. According to the New York Times in October 2014, Silicon Valley also has a diversity problem – one which is being tackled head on by companies such as Google and Facebook.

Academia has a lot to learn from how other kinds of large organisation have identified the nature and causes of gender inequality. On Wednesday 18 March the TRIGGER project and the BEI School are hosting a networking event designed to explore which institutional changes work best in supporting gender equality in large organisations. The panel’s speakers will reflect on why changes were necessary, what changes have been introduced, the outcome of those changes, and what still needs to happen to improve gender equality. The diversity of speakers will ensure there are opportunities for learning for all.

Interested? Find out more