Linda Flower & John R. Hayes’s 1981 Cognitive Process Model of composition serves as a base for evaluating composition activities, as Deborah Brandt notes (1992). Flower & Hayes identify three key cognitive elements of the writing process: the writer’s knowledge of topic, audience and context (also termed the “long-term memory”); the task environment (including everything external to the writer, the rhetorical problem, and the developing text); and the writing process itself (including planning activities, the actual writing of the text, and ongoing revision of the text). This model is a hierarchical model of composition, as opposed to a stage-based model: it describes the more fluid mental processes of composition, rather than a linear progression of activities from one stage to the next. For example, a writer is likely to engage in goal setting for their text at any point in the composition, reshaping the goals for the text as review of the produced text enhances the writer’s understanding of the rhetorical situation.
The model is not a perfect one, as it is largely self-contained to the particular text currently underway, and does not explicitly account for external influences such as interruptions, long-term breaks in the creation process, or simultaneous work on other texts. It is also notable that this cognitive process model does not in the first instance incorporate multimodal forms of creation, which Andy Campbell calls a “liquid canvas” (2011, n.p.); incorporating Flower & Hayes’ 1984 Multiple Representation Thesis, however, offers a more fluid aspect. This theory offers relevant insight into the development of Gunther Kress’s “synaesthetic process” (1998, in Fortune 2005, 53) necessary for multimodal composition: essentially, that it is already inherent in the process of composition. “Writers at work represent their current meaning to themselves in a variety of symbolic ways”, including nonverbal, procedural, and imagistic representations of ideas and knowledge (Flower & Hayes 1984, 129). The process of translating this abstract knowledge into written text is a difficult one, and the authors note that multimodal texts offer a significant advantage in that “some goals are better accomplished with different representations… Which representation is in force at a given moment is probably drven [sic] by a combination of one’s goals at that moment and the forms of the particular representation already stored in memory” (ibid., 151). The argument can be made here that composing multimodally engages more naturally and fluidly with the planning process of composition. Alan Sondheim (2006) and Jenny Weight (2006) respectively echo this thesis in their practice-based explorations of their own digital composition process, and Jason Ranker likewise describes this effect in his 2008 ethnographic study of students composing in digital media.
Embedding this Cognitive Process Model within the framework of Mihaly Csikszentmihalyi’s 1996 Systems Model of Creativity assists in consideration of external influences. The Systems Model defines creativity as occurring when “a person, using the symbols of a given domain…has a new idea or sees a new pattern, and when this novelty is selected by the appropriate field for inclusion into the relevant domain” (Csikszentmihalyi 1996, 28); this creative novelty either changes the domain, or transforms it to a new one. Domain encompasses a set of symbolic rules and procedures that identify an area of knowledge; field is the individuals who act as gatekeepers for that domain; person is used to identify the individual engaging in the creative activity, which Csikszentmihalyi notes requires an internalization of the system — familiarity with the domain and field in which the creative act is engaged. According to this model, an act, idea, or product is not creative unless it is acknowledged by the relevant domain and field (which can be difficult, depending upon the domain and field’s ability to recognize and incorporate the novelty’s validity and implications). Accepting that the person engaged in the act of composition employs Flower & Hayes’s “long-term memory”, and that this must, according to Csikszentmihalyi, incorporate knowledge of domain and field, offers a way to account for these external influences in the cognitive processes of composition.
Another gap in Flower & Hayes’s model rests in the “generating” box. Theirs is an encapsulated model of composition, offering a useful overview (at least from this writer’s perspective) of the major categories, but giving little attention to the age-old fan question: Where do ideas come from? The idea, “the concept that serves as the starting point for idea generation and elaboration, is potentially idiosyncratic for each writer and can be more or less elaborate, which will influence the richness of associations resulting from a memory search” (Lubart 2009, 151).
Incorporating The Geneplore Model (Finke, Ward, & Smith 1992 in Finke 1996) within the overarching framework of the “generating” phase of the Cognitive Process Model offers additional hierarchical levels of exploring the creative writing process. In this model, the authors propose a cycle of idea generation and exploration, which, like Flower & Hayes’s model, can be revisited as and when needed. The Geneplore Model’s generative processes mirror Flower & Hayes’s Multiple Representation Thesis (1984): “in addition to visualised patterns and object forms, [generative processes] may include mental blends, category exemplars, mental models, and verbal or conceptual combinations” (Finke 1996, 385). The generative process is a brainstorm of ideas pulling from existing examples, recombination of elements from those exemplars, and novel approaches to the rhetorical problem. The resulting preinventive structures can then be explored and interpreted, then reshaped as needed (per rhetorical situation, which includes product constraints) through further generative processes. For instance, this framework offered insight into how the cognitive effects of immersion in digital tools and environments led to fragmentation and layers of narration in my own work (Skains 2015).
Finally, an aspect of the composition process that should be incorporated is serendipity, defined as “a process of making a mental connection that has the potential to lead to a valuable outcome, projecting the value of the outcome and taking actions to exploit the connection, leading to a valuable outcome” (Makri & Blandford 2012b, 2, emphasis original). Arguably, serendipity is the confluence of cognitive activity and external stimulation that most often leads to so-called “eureka moments” for creators, and inspires the “how did you come up with such a great idea?” questions from audiences. S. Makri & Ann Blandford (2012a, 2012b) outline a model identifying this cognitive process as something more than luck; rather, it is the convergence of the knowledge and experience to make the mental connection and to recognize the significance of that connection, with the skills necessary to exploit the connection and produce a worthwhile outcome or artifact. Serendipity is likely behind the advent of many narrative evolutions, such as the combination of genres into new forms (tech-noir, space opera); the concept also enabled me to analyze the effects of digital appropriation in my multimodal fiction, digging deeply into how an idea developed and evolved through the processes of creation (Skains 2016).
I have gathered these cognitive and creativity models into a cohesive structure that best represents the composition context and cognition: the Practitioner Model of Creative Cognition. This model is based upon the strong foundation provided by Flower & Hayes’s Cognitive Process Model (1981), but widens it somewhat beyond the internal cognitive processes to incorporate the overall system of the practitioner’s creative context using Csikszentmihalyi’s Systems Model of Creativity (1996), allowing for the examination of external influences upon the writing process.
Makri & Blandford’s 2012 (a, b) model of serendipity is incorporated as a mediating function of the monitoring process, where expanding awareness of the domain, the field, and the emerging text converge to form an optimal state for serendipitous mental connections and discoveries. Within the generating process, I have embedded the Geneplore Model, in order to unpack the aspect of how ideas are shaped and remodeled (Ward, Smith, & Finke in Finke 1996). Flower & Hayes’s Multiple Representation Thesis (1984) offers insight into the translation process, whether the practice is mono- or multimodal. The translation process also now includes considerations of materiality; though materiality also clearly comes into play in the “product constraints” aspect of the Geneplore Model, it is a significant factor in the translation of narratives, particularly multimodal narratives.
Similarly, the fiction world has been embedded into the translation process as a distinct element, drawing from Todd Lubart’s argument that the writer in the process of translation is constantly shifting between the writing world and that of the fiction: “[t]he fiction world seems to involve productive thinking, improvisation, and a lack of reflective, evaluative thought...In contrast, the writer’s world is active, critical, and directive” (2009, 159). While consideration of the fiction world is inherent in monitoring, evaluating, and reviewing the text produced so far, there is also a specific aspect of translation in which the fiction world plays out independently from the writer’s goals and plans, and thus is worth additional consideration in the model.
This is a model formulated from introspection, self-observation, and reflection upon my own artistic practice, based upon the models discussed in this section; it has not been drawn from larger ethnomethodological studies of other practitioners at work. As such, it may be subject to future adjustments, and it may not be applicable to every individual. By drawing upon more widely accepted models, and integrating the insight of an experienced practitioner engaged in a targeted, practice-based project, however, the Practitioner Model of Creative Cognition gains validity.