Ill-Conceived Notions

An experiment in digital humanities

The first step of Grounded Theory is the open or initial coding phase. In this phase, we look for pointers on where the data can possibly take us, using a descriptive language that captures the actions and processes evident in the text. Put simply, initial codes question what the text is ‘doing’. Charmaz (2006) recommends using gerunds in each code, to “gain a strong sense of action and sequence ” (p.49), which can later tell us with more certainty what each category is ‘doing’ to represent a theme of the narrative.

In traditionally collected data, such as the examples offered by Charmaz of previously conducted interviews, a directed narrative has been prompted from a subject, rather in this case where a narrative is being offered of the speaker’s own accord, not with the intention of offering the reader insight to the speaker’s personal experience. This makes the coding of natural conversation slightly more tricky, since true sentiments and convictions are more implied than overtly explored. I performed multiple close readings, with even more holistic readings, of each dataset before finalising a language I felt described common themes of speech habits, which I could then organise into themes for each set. Charmaz describes this as a necessary process whereby the researcher asks themselves questions about what patterns are emerging, to inform them of what they need to look for as they go on. This is similar to other qualitative methods which apply a theory to be verified or use pre-existing coding schemas to strengthen pre-existing research, except in this case, every idea is emerging solely from the data. The initial coding process, as Charmaz describes, is quick and and intensely descriptive, to capture the first interpretation we make of the data. This provides a rich description from which we can build a multitude of potential theories.

To examine the language I was using in the initial coding, I used the ‘Word Frequency Query’ and noted each gerund I used in an Excel file. For Molyneux’s videos, my initial codes utilised 97 gerunds for ‘Crusades’, 39 for ‘Enslavement’, and 58 for ‘Mandela’. For each, my most used words were ‘emphasising’, ‘implying’ and ‘framing’ (in different orders) and I noticed that many of the words I used were referring to similar ideas. For example, ‘expressing’ could cover ‘sympathising’, ‘desiring’ and ‘wishing’ (each of which I only used once or twice) if I amended my phrasing of those codes, and it more accurately alluded to an overarching theme of implied personal connections with the content. I went through each video a second time, noting each gerund as I used it, and found I could achieve the same meaning with a much more concise language, ending up with 28 gerunds total for the 3 videos after my final review of the codes. I wrote descriptions for each of these to ensure the meaning in each instance was consistent and I was not using multiple words to capture the same idea. For example, each instance where I explained Molyneux was ‘emphasising’ something, I meant he was “using some tactic to make a certain idea or topic stand out in a sentence, such as shouting, repeating, or speaking slowly.” Similarly, for the ContraPoints videos, I initially utilised 73 gerunds for ‘Incels’ and 48 for ‘Peterson’, which I refined to 35 for both videos. This made it much easier to see the recurring themes in the videos, and patterns that I needed to look out for in follow-up readings as I continued to add details that my narrowed my focus.

A leading note I found coding the ContraPoints videos were changes in tone, since she could switch from inquisitive to sarcastic to thoughtful in a very short time. I believe I noticed it more so with her videos because her tone is steady and neutral by default, and her shifts in tone are much more obvious since they are rehearsed. After finishing the ContraPoints videos I revisited Molyneux’s to add notes on the changes in tone, which I had skimmed over not realising the large component it played in changing the intention of his words. I had a few notes on this already, but on my next round of coding I listened exclusively to the tone and flow he used for each segment since I had coded for every other action in his speech, as I had for ContraPoints. ‘Enslavement’ especially was loaded with changes in tonality to emphasise certain aspects of the message, such as Molyneux’s submissive offering of indisputable truths of life, to his irritation when framing governmental oppression, with fluctuations between harsh and soft tones to guide the emotional response of the audience. I added adverbs and adjectives to the lists of gerunds I had made for Molyneux and ContraPoints, also adding descriptions to help me see where and how I was using each word.

My run-through of the focused coding for the last ContraPoints video was easily the fastest, since by then I had clear ideas already of what it was I was looking for, and virtually all the descriptions written for the gerunds and adjectives. While I still aimed to highlight any new ideas that may not have appeared in the other videos, the focused coding was more a matter of deciding what details I needed to add to capture tones or notes I may have missed the first time.

Leave a Reply

Your email address will not be published.