Up: 011_data-and-disc-lect Prev: week13-corpus-approaches-to-discourse-1
Video Lecture Notes:
Understanding Corpus Approaches to Discourse: A City Analogy The lecture begins with an illuminating analogy comparing discourse analysis to exploring a city. Just as understanding a city requires both close-up inspection and a bird’s-eye view, discourse analysis demands both detailed textual examination and broader statistical analysis. This dual approach helps researchers understand both specific elements and larger patterns within discourse.
Defining Corpus and Its Characteristics A corpus is defined as a collection of texts containing natural language utterances produced in authentic situations. The lecture identifies several key types of corpora:
- Specialized Corpora: Collections assembled for specific studies
- Reference Corpora: Broader collections compiled for entire research fields (e.g., the British National Corpus)
The fundamental characteristics of a corpus include:
- Machine-readable content
- Authentic texts and transcripts
- Representative sampling of language or language varieties
The Challenge of Representativeness An important theoretical consideration in corpus linguistics is the concept of representativeness. The lecture explains that achieving true representativeness is challenging because the total population of discourse (all words uttered within a discourse) is usually unknown. To address this limitation, three approaches are presented:
- Large Versatile Corpus: Maximizes size and allows stratification
- Balanced Corpus: Includes diverse genres, sources, and media
- Specialized Corpus: Focuses on specific discourse sections or topics
The Layers of Discourse Analysis The lecture outlines four interconnected layers when investigating discourses with corpora:
- Linguistic Expressions: The form side of words and phrases
- Concepts: The meaning side of linguistic signs
- Linguistic Practices: Actions performed through language
- Social Practices: Real-world activities that leave linguistic traces
Practical Applications and Tools The lecture describes several practical aspects of corpus research:
- Corpus Workbench: A research environment featuring:
- Indexed database of words
- Corpus Query Language (CQL)
- Web interface (CQPweb)
- Analysis Techniques:
- Concordance Analysis: Examining words in their linguistic context
- Collocation Analysis: Measuring statistically significant word associations
- Keyword Analysis: Identifying significant vocabulary by comparing frequencies
The Role of Context and Meaning The lecture emphasizes John Rupert Firth’s principle: “You shall know a word by the company it keeps.” This underscores that meaning in corpus linguistics must be reconstructed through:
- Analysis of word patterns
- Understanding contextual relationships
- Examination of linguistic configurations
Interdisciplinary Nature The lecture concludes by highlighting the importance of interdisciplinarity in discourse analysis. While corpus approaches can reveal patterns in expressions, concepts, and linguistic practices, understanding social practices requires collaboration across disciplines.
Practical Example: Climate Change Discourse The lecture uses climate change discourse to illustrate how corpus analysis works in practice:
- Examining expressions like “climate change” or “cambio climatico”
- Analyzing linguistic practices (warnings, descriptions, denials)
- Understanding how these connect to social practices (law changes, emission reductions)
This comprehensive approach to discourse analysis demonstrates how combining quantitative corpus analysis with qualitative interpretation can provide deeper insights into language use and social meaning. The methodology continues to evolve, particularly with the growth of digital discourse analysis and dedicated journals in the field.