Language as an Embodied and Enactive Process: What Underlies Expressive Writing?
Human language is a fascinating aspect of humankind and serves different purposes like expressing ourselves, reasoning, or transmitting knowledge over generations; all in all, it is not surprising how it has raised interest for decades in research.
Expressing oneself through language, either in speech or in writing, is a complex performative act involving interrelated processes. This is true when we are in a heated argumentation with a colleague about an important issue, when we are in a stimulating conversation with a friend, or when we choose our words carefully when writing a personal letter. In all these situations, we are not merely sending messages across. We are expressing ourselves through body and mind. And the words that come from our mouth, our pen, or keyboard are not merely the end products of a linguistic mental process. Rather, they reflect a stream of closely coordinated perceptions, movements, thoughts, and emotions, observable by others through listening or reading. In other words: We are linguistic bodies (Di Paolo, Cuffari and De Jaegher, 2018), which is exemplified further by the many nonverbal aspects of communication.
The concept of linguistic bodies (Di Paolo, 2018) may be unfamiliar, but becomes immediately clear when you pay attention to a person’s eye movements, facial expressions, gestures, and the pitch and intonation of voice, during ‘in real life’ speech. All these things, as well as what is actually said, are obviously related to the speaker’s cognitive and emotional state. In fact, understanding natural language is only possible because we are able to tune into the dynamics of everything that is going on, and not just the words that are being spoken. That similar processes are at play during writing and reading probably is a bit more surprising. But as we shall see, even a text reveals a great deal more than just the message contained in the sequence of written words.
So what is it that words tell us beyond what is written? And how do they reflect the interplay of many things that are going on within and around a person? Moreover, can we, as researchers, use this to better understand people’s minds or measure their wellbeing? In an attempt to expose our linguistic bodies, this post will introduce how it’s possible to analyze expressive writings and how this can provide insight into personality and mental health. For this, techniques from the toolbox of complex dynamical systems theory will be integrated.
The research ideas presented here are part of the PhD project of Nicol Arellano Véliz on embodied markers and intersubjectivity of psychological traits, supervised by dr. Ralf Cox.
Linguistic Bodies as continuity between life and language: A starting point
Language operates in continuous interaction with our environment. It cannot uniquely operate inside our minds, we learn it and use it in social contexts. Language is not merely a reflection of our inner state of being, but a fundamental complement to it (Clark and Chalmers, 1998; Clark, 2008). For instance, a poet while writing is expressing thoughts and feelings, labeling corporeal sensations as emotions, and transforming personal perceptual experiences of color, sound, smell, or images, at the same time —during the act of writing— is engaged in perception-action cycles. We can see that language is a constant interplay between different constraining influences, verbal and nonverbal, in addition to bodily sensations and subjective experiences. Body, brain and environment operate as wholeness in creating meaning, and language is shaped by our embodied selves, involving sensorimotor and affective processes (Johnson, 2018).
Earlier, when I said that we are linguistic bodies, I was referring to the organic, sensorimotor, intersubjective, and linguistic dimensions of our bodies, as understood by researchers from the enactivist approach (Di Paolo, 2018; 2020). Enaction refers to the continuing process of being coupled to the environment by sensorimotor activity, which is only possible by the organic (i.e. biological) dimension (Varela, Thompson & Rosch, 2016; Thompson, 2017). This may seem obvious, but we need a body to perceive and integrate all the sensory information, leading to a behavioral or affective response. Then, language is enactive because it emerges in a continuous interactional way. Consequently, the intersubjective dimension implies that language originally arises from interactions with others.
However, in principle, there is no need for another person to be physically present. Language can be used for individual purposes, like developing and expressing thoughts or feelings in solitude through writing. For this, we can make use of external objects to transmit or deposit information. In those cases, language takes place in interactions with devices like laptops, smartphones, or the old-fashion paper and pencil, which serve as extensions of our bodies and minds (Johnson, 2018). This form of language expression is especially relevant for the research we will present next.
Body, brain and environment operate as wholeness in creating meaning, and language is shaped by our embodied selves, involving sensorimotor and affective processes.
Human Language as a Complex System: The case of expressive writing
Now that the idea of linguistic bodies was introduced, I will talk about written language, particularly, its expressive form —in which our research is focused. Expressive writing is referring to the act of writing freely, for example when you are writing a personal letter or journaling, organizing your thoughts and personal experience in a text. An expressive writing piece can be analyzed and give us information about psychological processes like personality and mental health, by using the methods of the complex dynamical systems.
Fundamental to our approach are the underlying dynamic patterns in written texts; the sensorimotor, cognitive, and affective processes reflected by them. It is not only the content and explicit meaning but also the ordered structure of that corpus of written words that contains information. Thus, for example, a text like this blog [or anything else] can be analyzed by the meaning of the words, but also by looking at the length and repetition of the words. Fundamentally, written human language is a dynamical and complex system that involves many units differing in function, length, and frequency (Rodriguez et al., 2014).
What are the implications of understanding language as a complex dynamical system?
Basically, it implies that its functioning and expression are composed of different elements and relations between them (quantifiable), either letters, words, sentences, or paragraphs. We can see material patterns (e.g. types of words and sentence structure) enacted by us as linguistic bodies in a deterministic sequential order, which is the materialization of the person’s cognitive and affective’s operations (Di Paolo, 2018).
To analyze this, the texts made by the participants of our study are conceived like structured temporal sequences of events, as a time series so to speak, reflecting the dynamics of the linguistic body’s writing activity. This implies reducing the multidimensional (and abstract) dimension of thoughts into a one-dimensional stream of letters, words, sentences, and paragraphs (Chatzigeorgiou et al. 2017). In addition to analyzing the text’s content, we can open up the toolbox of complexity science, which allows us to detect and quantify long-range correlations, patterns, and symmetries in these sequences (Chatzigeorgiou et al. 2016; Skovgaard et al. 2019). And all these elements are indicative of the very inner structure and dynamic organization of that text.
What kind of information can expressive writing give us?
This approach gives us a different and cohesive scope of measurement. In this case, our thoughts are shaped as correlations in the time series throughout the text. For instance, the repetition of words in the text composition and the length of the words and sentences can help us to detect its inner coherence (Najafi & Darooneh, 2017). On the other hand, through content analysis, ranking the importance of the words, and performing sentiment analysis —which scores the text’s emotionality according to its affect valence— we are trying to explore the connections between these material patterns and “higher-order” psychological concepts such as Personality and Depression.
Even if we are studying the structure of a text and quantifying indicators extracted from it, we are assuming the existence of interrelated processes involved in the act of writing beyond that quantification. We assume that the ways in which someone experiences life will make an imprint in the sequence of words while writing. Thus, the dynamic information extracted from the text itself should give us information about the person who wrote it, going beyond the words themselves. This is because writing emerges from a collection of sensorimotor, cognitive, and emotional constraints in combination with syntactic and semantic operations, and that process is colored by the idiosyncrasies of the individual writer.
Then, what do we expect to find?
We expect to find connections between the text structure and psychological constructs, that is, differential patterns in the inner structure of the texts for different individuals. We aim to detect specific correlates for each personality dimension and also “typologies” (combinations of personality dimensions). For example, differentiated trajectories of emotionality (valence) among the narratives. Perhaps a person with high scores in extraversion and emotional stability will tend to show stable and positive emotional content through the text. Also, differences in the type of words used are expected, for instance, a person who is open to experience will tend to use less common words. In addition, the texts of such a person might be less deterministic and structured. In the same way, we believe that it is possible to detect healthy and unhealthy patterns in the system’s behavior (in this case, a person). In this case, we expect to be able to differentiate texts written by individuals who report depressive symptoms and those who don’t, based on particular markers. This idea can also be extended to other psychological constructs.
We believe that these methods can contribute to the development of additional forms of assessments of psychological phenomena and further multimodal tools of diagnosis for clinical conditions to triangulate self-report and clinical criteria. To conclude, we expect to obtain key information from the linguistic bodies which goes beyond the explicit elements contained in a text, as a way to access high-order processes of human experience. Like Giampiero Arciero (2009) explains in a very clever way:
Through language when the ‘itself’ opens up, we can communicate our own experiences of existing, and at the same time, they are being constantly shaped (and changing) by the agency of language.
Note: This post was written with the collaboration and feedback of dr. Ralf Cox.
Featured Image: Creative Commons CC0 1.0 Universal Public Domain Dedication
Arciero, G. & Bondolfi, G. (2009). Selfhood, Identity and Personality Styles. West Sussex, UK: John Wiley & Sons.
Chatzigeorgiou, M., Constantoudis, V., Diajonos, F., Karamanos, K., Papadimitriou, C., Kalimeri, M. & Papageorgiou, H. (2016). Multifractal correlations in natural language written texts: Effect of language family and long word statistics. Physica, 469, 173-182. doi: 10.1016/j.physa.2016.11.028
Clark, A., & Chalmers, D. (1998). The Extended Mind. Analysis, 58(1), 7-19. doi: 10.1111/1467-8284.00096
Clark, A., (2008). Supersizing the mind: embodiment, action, and cognitive extension. New York, NY: Oxford University Press.
Di Paolo, E., Buhrmann, T. & Barandiaran, X. (2017). Sensorimotor life. An Enactive proposal. Oxford: Oxford University Press.
Di Paolo, E., Cuffari, E., & De Jaegher, H. (2018). Linguistic Bodies. The Continuity between Life and Language. Cambridge, Massachusetts: The MIT Press.
Di Paolo, E. (2020). Enactive becoming. Phenomenology and the Cognitive Sciences. doi: 10.1007/s11097-019-09654-1
Johnson. M. (2018). The Embodiment of Language. In A. Newen, L. De Bruin, & S. Gallagher (Eds.). The Oxford Handbook of 4E Cognition. doi: 10.1093/oxfordhb/9780198735410.013.33
Najafi, E., & Darooneh, A. (2015). The fractal patterns of words in a text: A method for automatic keyword extraction. PLoS ONE, 10(6), 1-18. doi: 10.1371/journal.pone.0130617
Najafi, E., & Darooneh, A. (2017). Long-range dependence in texts: A method for quantifying coherence of text. Knowledge-Based Systems, 133, 33-42. doi: 10.1016/j.knosys.2017.06.032
Skovgaard, M., Mehlsen, M., Bonde, J., Howard, D., Philipse, J. & Wallot, S. (2019). Use of Recurrence Quantification Analysis to examine association s between changes in text structure across an expressive writing intervention and reductions in distress symptoms in women with breast cancer. Frontiers in Applied Mathematics and Statistics, 5(37), 1-13. doi: 10.3389/fams.2019.00037
Thompson, E. (2017, January 27). The Enactive Approach [Blog post]. Retrieved from https://philosophyofbrains.com/2017/01/27/the-enactive-approach.aspx
Varela, F. J., Thompson, E., & Rosch, E. (2016). The embodied mind: Cognitive science and human experience. (Rev.ed.). Cambridge, MA: The MIT Press.