The interplay of complexity and subjectivity in opinionated discourse
Bibliographic record
Abstract
This paper brings together cutting-edge, quantitative corpus methodologies and discourse analysis to explore the relationship between text complexity and subjectivity as descriptive features of opinionated language. We are specifically interested in how text complexity and markers of subjectivity and argumentation interact in opinionated discourse. Our contributions include the marriage of quantitative approaches to text complexity with corpus linguistic methods for the study of subjectivity, in addition to large-scale analyses of evaluative discourse. As our corpus, we use the Simon Fraser Opinion and Comments Corpus (SOCC), which comprises approximately 10,000 opinion articles and the corresponding reader comments from the Canadian online newspaper The Globe and Mail, as well as a parallel corpus of hard news articles also sampled from The Globe and Mail. Methodologically, we combine conditional inference trees with the analysis of random forests, an ensemble learning technique, to investigate the interplay between text complexity and subjectivity. Text complexity is defined in terms of Kolmogorov complexity, that is, the complexity of a text is measured based on its description length. In this approach, texts which can be described more efficiently are considered to be linguistically less complex. Thus, Kolmogorov complexity is a measure of structural surface redundancy. Our take on subjectivity is inspired by research in evaluative language, stance and Appraisal and defined as the expression of evaluation and opinion in language. Drawing on a sentiment analysis lexicon and the literature on stance markers, a custom set of subjectivity and argumentation markers is created. The results show that complexity can be a powerful tool in the classification of text into different text types, and that stance adverbials serve as distinctive features of subjectivity in online news comments.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.014 | 0.080 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.007 | 0.005 |
| Science and technology studies | 0.002 | 0.009 |
| Scholarly communication | 0.007 | 0.012 |
| Open science | 0.001 | 0.004 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".