On March 07, 08, and 09, Helen and Djuddah attended the 2022 online International Researchers Consortium (IRC) workshop as authors of our project and readers of other projects around the globe. The IRC workshops are organised annually by the Conference on College Composition & Communication. The IRC workshop has been an integral part of our research work in the past, and the feedback we receive on the project has always been valuable. This year, we received feedback from Christiane Donahue, Julie Koligjini, Adrienne Lamberti, on our workshop proposal, where we presented our feature model and asked for feedback on the following questions:
-
- Throughout the project, the terminology has been an aspect we’ve been challenged by and are challenging. Terms such as Writing Tradition, Writing Culture, Writing Structure – how do you interpret these terms in your own context?;
- One of our main aims of the project is to develop a model which will help us to measure (deduce the component that makes an academic text) academic texts. The model is not meant to be a fixed model, but a model which shifts fluidly to accommodate for differences (genres, disciplines, language), we wonder whether you think that all necessary features are represented in the model itself, whether you think any features or sub-features are missing, or whether you would delete these features (considering different languages). In other words, how coherent is the model as a unit and how well do you think it represents a whole academic text?
- Our methodological approach to measuring academic texts is using machine learning algorithms. One of our aims in this project is to develop a method that will allow us to use machine learning methods to run predictive algorithms on a large dataset of academic texts, which should highlight how these texts cluster or fit in a specific model. To achieve this stage, we first need to identify aspects/features that can be measured in a text (for example coherence, anaphoric relations, citation patterns, etc.) and identify what these features say about a text in a specific context (e.g. citation patterns can say something about argumentation). Our questions about methodology are: how do you evaluate this approach to modelling in general, but also to build a coherent understanding of a writing tradition, culture, writing structure? And, what kind of results would you assume one would find with such an approach (in your own context?); and
- In our draft, we have identified the key theorists guiding our approach. In addition to these, do you have any suggestions of theoretical approaches we are missing?
The main takeaway we took from our workshop was: we received some very useful suggestions about features we could further include or perhaps further investigate as possible features. One specific is the inclusion of metaphor. How to operationalise metaphor is something we’ll have to look into. We received a useful suggestion to look into the handbook They Say / I Say written by Gerald Graff and Cathy Birkenstein as possible means to code for citation patterns. Other suggestions relate to ways we can conceptualise the processes which take place in Estonia, Latvia, and Lithuania regarding writing and genres being disciplined or the acculturation of specific writing traditions.
