Cognitive Discourse Analysis (CODA): Finding out how people think by looking at what they say (and how they say it)

Now published as a textbook: Tenbrink, Thora (2020). Cognitive Discourse Analysis: An Introduction. Cambridge: Cambridge University Press.

Thora Tenbrink, January 2020

How can we learn about what goes on in people’s minds? Cognitive scientists have considered this for a long time, and have developed many answers, a wide range of very useful methods. However, not many of them are as easy to access as language. We ask people what they are thinking - and they will answer in language. It’s not as simple as that, of course: What people say will not be exactly the same as what they are thinking. Language is only a representation of thought, but it is a powerful and very flexible and informative representation. CODA (Cognitive Discourse Analysis) offers a systematic way of making sense of this kind of representation - a way of learning about the human mind by taking a close look at what people say and how they say it.

Several linguistic theories address the relationship between language and thought (e.g., Talmy 2000, 2007; Evans and Green, 2006; Langacker, 2000; Tomasello, 2003). Various kinds of structures in language appear to be systematically related to cognitive structures and processes. This principle of language as a system also carries over to how this system of language is used: the way we think is related to the way we talk. This is true both generally in terms of what we can do with language, and specifically with respect to what we actually do.

The relationship between language and thought may not be simple, but it is nevertheless systematic. For instance, if you look at a scene and describe what you see in the scene, your description will reflect what you’ve noticed in the scene - where your attention was, and in what order you were considering and describing aspects of the scene. There will be some aspects that you don’t mention at all, and some others which you only mention in passing rather than elaborating them in detail. This indicates the sequential processes of attention - as far as they are represented in language.

In route directions, an utterance like “Turn right at the shopping mall” shows that the speaker has a concept of a unique shopping mall that distinguishes it from other buildings in the environment, and can therefore be referred to by a definite article and used as a landmark to anchor a direction change. In the context of mountaineering, such a concept doesn’t work so well: people will then resort to environmental features such as “where the snow begins” as landmarks (Egorova et al., 2015). To take another example, if speakers orient towards the overall trajectory of a planned path, or focus on object features such as colour to make their decisions, these tendencies will be reflected in their linguistic choices of references to path or object features (Tenbrink & Wiener, 2009). In these and other ways, linguistic choices reflect crucial aspects about the speakers’ concepts and strategies at any given moment. This provides a good pathway to access cognition, given the necessary expertise about relevant features of language.

When asked to verbalise their thoughts, as in these examples, speakers make a lot of choices in a very short amount of time (Halliday & Matthiessen, 2014). The general repertory of a language is vast: an infinite range of things can be expressed in an infinite range of different ways. People choose from this repertory when they speak, and they do so in systematic ways - depending on the nature of their current thoughts, and the overall context of the situation. A systematic analysis of these choices in relation to a situation therefore reveals fundamental aspects about how people thought of the situation. For instance, when people think aloud while they’re solving a problem, their thought processes during problem solving can be identified (Ericsson & Simon, 1994). And when people describe something they remember, for instance an event or a film, their formulations can shed light on specific aspects of their memory.

Beyond content analysis, grammatical theory, cognitive linguistic semantics, and other linguistic findings support the interpretation of language use in relation to cognition. For instance, if somebody says “Our journey together has been on bumpy roads at times, but now we’ve been married for 10 years and are still going strong”, the analyst will easily recognise a well-researched conceptual metaphor where LOVE is thought of and described as a JOURNEY (“journey together”; “bumpy roads”; “going”) (Lakoff & Johnson, 1980). The fact that the speaker uses this metaphor reflects its relevance at this moment of speaking. If many people use similar metaphors in the same situation, but they are rarely used in another situation, this may reveal interesting patterns concerning the way people think about these different situations.

Along these lines, CODA uses linguistic insights and analysis procedures in order to find out more about how people think. The method was primarily designed for the analysis of unrestricted language production data collected in controlled studies that relate to research issues in cognitive science, following up on ideas presented in Caron-Pargue & Gillis (1996). In this field, questions about human concepts and thought processes are often addressed by using various kinds of unconstrained verbalisations. CODA provides a framework for utilising linguistic insights for the analysis of such data, by investigating patterns of language use in relation to the situation in which language is produced. Relevant studies involve situations or tasks that highlight central aspects of mental representation (the conceptualisation of complex scenes, event perception, and the like) and complex cognitive processes (such as problem solving or decision making). Both of these relate to and enhance well-established research traditions in distinct ways.

With respect to mental representation, CODA addresses the conceptualisation of perceived situations and events, building on established psycholinguistic methods (e.g., Ellis, 1985/1987). In comparison to conventional psycholinguistic studies, the experimental settings used in CODA are less constrained so as to allow for free language usage, leading to a more thorough integration of qualitative (descriptive) findings concerning patterns in unconstrained linguistic choices.

For example, in a collaboration with Kenny Coventry and Elena Andonova we carried out a series of empirical studies addressing spatial concepts and description strategies of spatial configurations. Results revealed that different object arrangements affected description strategies across various linguistic levels, including description trajectories, amount of detail, and explicit mention of atypical object orientation (Andonova et al., 2010). A detailed analysis of description strategies highlighted patterns in the use of conceptual reference frames and perspectives (Tenbrink et al., 2011). Complementary insights on spatial description strategies in more abstract settings were found by Tenbrink & Ragni (2012). In contrast to these structured scenarios, description strategies in cluttered environments were far more varied and incomplete (Tenbrink, Bergmann, Hertzberg, & Gondorf, 2016).

Other CODA studies highlight principles of mental representation across diverse contexts. For instance, Mast et al. (2014) identified two distinct principles of categorising direction in speakers’ spatial descriptions, shedding new light on the relationship between linguistic and non-linguistic categories. Danino (2014) used CODA to analyse meaning construction in the discourse of 9/11. Egorova et al. (2018) investigated expressions such as “our route traversed much further to the right” in alpine narratives, showing how speakers make use of ‘fictive motion’ (it is not the route that traverses!) to express their movements. Thomas et al. (2015) found interrelations between linguistic markers of hesitation and accuracy in eyewitness reports. And Cialone et al. (2018) showed how professional background systematically affects how people perceive and describe pictures.

With respect to complex cognitive processes, CODA enhances the widely used research paradigm of using think-aloud protocols and retrospective reports for the identification of (internal) cognitive processes (Ericsson and Simon, 1984). In that paradigm, researchers typically focus on the content of verbal data, predominantly addressing those aspects (e.g., particular thought processes or strategies) that the speakers are themselves aware of. The content-based inspection of verbal reports, particularly if carried out by experts in the problem domain and set against a substantial theoretical background (Krippendorff, 2004), can lead to substantial hypotheses about the cognitive processes involved. The systematic analysis of linguistic structures (as in CODA) then provides a sound basis for using the language data as evidence, based on a systematic operationalisation motivated by the relevant literature on linguistic features. Furthermore, it may lead to additional important discoveries concerning human conceptualisation that may not be accessible by employing content-based analysis uninformed by linguistic expertise.

In CODA, verbal reports are analysed with respect to patterns of linguistic structures emerging across individual speakers, constrained by their strategies of dealing with the problem solving task at hand. Relevant scenarios encompass, for instance, analogical problem solving (Gralla, Tenbrink, Siebers, & Schmid, 2012), planning and description strategies in a complex building (Tenbrink, Bergmann, & Konieczny, 2011), conceptual strategies in the Travelling Salesperson Problem (Tenbrink & Wiener, 2009; Tenbrink, 2008), holiday route planning (Tenbrink & Seifert, 2011), Origami paper folding (Taylor & Tenbrink, 2013, Tenbrink & Taylor, 2015), and maintaining orientation in virtual reality (Tenbrink & Salwiczek, 2016).

In both areas, mental representation and problem solving, the aim in CODA is to gain insights into generalisable cognitive phenomena that go beyond conscious reflection by individual speakers, and that may not necessarily be directly observable in linguistic content. Discourse participants may not be aware of the cognitive structures that are reflected in particular ways of framing a representation linguistically. Furthermore, they may not be aware of the underlying network of options (Tenbrink & Freksa, 2009) that allows for a range of linguistic choices beside their own, which emerges only by considering a larger data set collected under controlled circumstances. According to previous research in cognitive linguistics and discourse analysis (e.g., van Dijk, 2008), linguistic features such as the verbal representation of semantic domains reflected in ideational networks, lexical omissions and elaboration, presuppositions, hesitation and discourse markers, and the like all indicate certain conceptual circumstances; these should be systematically related to the mental representations or problem solving procedure in question. In addition, the relationship between linguistic patterns and the associated cognitive processes can be fruitfully examined further via triangulation, i.e., the combination of linguistic analysis with other types of evidence such as behavioral performance data (e.g., Tenbrink & Wiener, 2009; Hoelscher et al., 2011).


  • Andonova, Elena, Thora Tenbrink, and Kenny R. Coventry. 2010. Function and Context Affect Spatial Information Packaging at Multiple Levels. Psychonomic Bulletin & Review 17, 575-580.

  • Caron-Pargue, Josiane and Steven Gillis (eds), 1996. Verbal Production and Problem Solving. Antwerp Papers in Linguistics 85, Universiteit Antwerpen

  • Cialone, Claudia, Tenbrink, Thora, and Spiers, Hugo J. 2018. Sculptors, architects, and painters conceive of depicted spaces differently. Cognitive Science 42, 524–553.

  • Danino, Charlotte. 2014. Language production and meaning construction mechanisms in the discourse on an ongoing event: the case study of CNN’s live broadcast on 9/11. PhD Dissertation, University of Poitiers, France.

  • Egorova, Ekaterina, Tenbrink, Thora, and Purves, Ross. 2015. Where snow is a landmark: Route direction elements in alpine contexts. In Sara Irina Fabrikant, Martin Raubal, Michela Bertolotto, Clare Davies, Scott Freundschuh, and Scott Bell (eds.), Spatial Information Theory, pp. 175-195. Berlin: Springer.

  • Ellis, Andrew W. 1985, 1987. Progress in the Psychology of Language, 3 vols. London: Lawrence Erlbaum.

  • Ericsson, K. Anders and Herbert A. Simon. 1984. Protocol analysis: Verbal reports as data. Cambridge, MA: Bradford books/MIT Press.

  • Evans, Vyvyan and Melanie Green. 2006. Cognitive Linguistics. An Introduction. Edinburgh: Edinburgh University Press.

  • Gralla, Linn, Thora Tenbrink, Michael Siebers, and Ute Schmid. 2012. Analogical problem solving: Insights from verbal reports. In Naomi Miyake, David Peebles, and Richard P. Cooper (eds.), Proceedings of the 34th Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society, pp 396-401.

  • Halliday, M.A.K. and Christian M.I.M. Matthiessen. 2014. Halliday’s Introduction to Functional Grammar (4th edition). London: Routledge.

  • Hoelscher, Christoph, Thora Tenbrink, and Jan Wiener. 2011. Would you follow your own route description? Cognition 121, 228-247.

  • Krippendorff, Klaus. 2004. Content Analysis: An Introduction to its Methodology (2nd ed.). London: Sage.

  • Lakoff, George and Mark Johnson. 1980. Metaphors we live by. Chicago, IL: University of Chicago.

  • Langacker, Ronald W. 2000. A dynamic usage-based model. In M. Barlow and S. Kemmer (eds.), Usage-based models of language. Stanford, CA: CSLI, pp. 1-64.

  • Mast, Vivien, Wolter, Diedrich, Klippel, Alexander, Wallgrün, Jan Oliver, and Tenbrink, Thora. 2014. Boundaries and Prototypes in Categorizing Direction. In Christian Freksa, Bernhard Nebel, Mary Hegarty, and Thomas Barkowsky (Eds.), Spatial Cognition 2014, Bremen, Germany, 15-19 September 2014, pp. 92–107. Berlin: Springer.

  • Talmy, Leonard. 2000. Toward a Cognitive Semantics, 2 vols. Cambridge, MA: MIT Press.

  • Talmy, Leonard. 2007. Attention phenomena. In D. Geeraerts & H. Cuyckens (eds.), Handbook of Cognitive Linguistics. Oxford: Oxford University Press, pp. 264-293.

  • Taylor, Holly and Thora Tenbrink. 2013. The spatial thinking of Origami: Evidence from think-aloud protocols. Cognitive Processing 14:189–191.

  • Tenbrink, Thora. 2015. Cognitive Discourse Analysis: Accessing cognitive representations and processes through language data.Language and Cognition 7:1, 98 – 137.

  • Tenbrink, Thora. 2008. The verbalization of cognitive processes: Thinking-aloud data and retrospective reports. In Wiebke Ramm & Cathrine Fabricius-Hansen (eds.), Linearisation and Segmentation in Discourse. Multidisciplinary Approaches to Discourse 2008 (MAD 08), Feb 20-23 2008, Lysebu, Oslo. Oslo: Dept. of Literature, Area Studies and Europ. Languages, Univ. of Oslo, pp. 125-135.

  • Tenbrink, Thora, Bergmann, Evelyn, Hertzberg, Christoph, and Gondorf, Carsten. 2016. Time will not help unskilled observers to understand a cluttered spatial scene. Spatial Cognition and Computation 16:3, 192-219.

  • Tenbrink, Thora, Evelyn Bergmann, and Lars Konieczny. 2011. Wayfinding and description strategies in an unfamiliar complex building. In Laura Carlson, Christoph Hoelscher, and Thomas F. Shipley (Eds.), Proceedings of the 33rd Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society, pp 1262-1267.

  • Tenbrink, Thora, Kenny R. Coventry, and Elena Andonova. 2011. Spatial strategies in the description of complex configurations. Discourse Processes 48:237–266.

  • Tenbrink, Thora, Tommaso D’Odorico, Christoph Hertzberg, Guezin Mazman, Chiara Meneghetti, Nina Reshoeft, Jinlong Yang. 2012. Tutorial report: Understanding spatial thought through language use. JOSIS Journal of Spatial Information Science 5:107-114.

  • Tenbrink, Thora and Christian Freksa. 2009. Contrast sets in spatial and temporal language. Cognitive Processing 10 Supplement 2, S322-S324.

  • Tenbrink, Thora and Marco Ragni. 2012. Linguistic principles for spatial relational reasoning. In Cyrill Stachniss, Kerstin Schill, and D. Uttal (Eds.): Spatial Cognition 2012, LNAI 7463, pp 279-298. Springer, Heidelberg.

  • Tenbrink, Thora and Lucie Salwiczek. 2016. Orientation and metacognition in virtual space. Journal of Experimental Psychology: Human Perception and Performance 42(5), 683-705.

  • Tenbrink, Thora and Inessa Seifert. 2011. Conceptual Layers and Strategies in Tour Planning. Cognitive Processing 12:1, 109–125.

  • Tenbrink, Thora and Holly A. Taylor. 2015. Conceptual transformation and cognitive processes in Origami paper folding. Journal of Problem Solving 8:1, 2-22.

  • Tenbrink, Thora and Jan Wiener. 2009. The verbalization of multiple strategies in a variant of the traveling salesperson problem. Cognitive Processing 10:2, 143-161.

  • Thomas, Ayanna, Chen, Caroline, Gordon, Leamarie, and Tenbrink, Thora. 2015. Choose your words wisely: What verbal hesitation indicates about eyewitness accuracy. Applied Cognitive Psychology 29:5, 735–741. doi: 10.1002/acp.3157.

  • Tomasello, Michael. 2003. Constructing a Language: A Usage-Based Theory of Language Acquisition. Cambridge, MA: Harvard University Press.

  • van Dijk, Teun A. 2008. Discourse and Context. A sociocognitive approach. Cambridge: Cambridge University Press.