Notes

Conceptual change is introduced as an idea relevant for understanding learning. This is referenced as corresponding to *representational* changes. Representation in this context corresponds to the mental representation of concepts. This means that here, representation is more closely aligned with the establishment of a model (that is, the representation is the model in one’s head), as opposed to a physical representation of the model itself. This distinction is one that is important in discussing the relationship between mental modeling and simulation.

The challenge with articulating conceptual change is understanding the nature of how concepts are held mentally, and the mechanisms by which they are changed. The distinction given is between scientists and children, who are different in the sophistication of their goals and metacognitive strategies. Nercessian describes problem solving as a means by which to articulate how both groups might revise their conceptual models.

Describing concepts: “Concepts provide a means through which humans make sense of the world. In categorizing experiences we sort phenomena, noting relationships, differences, and interconnections among them. A conceptual structure is a way of systematizing, of putting concepts in relation to one another in at least a semi – or locally – coherent manner.”

On mental models: “Loosely construed, a model is a representation of a system with interactive parts with representations of those interactions. Models are representations of objects, processes, or events that capture structural, behavioral, or functional relations significant to understanding these interactions. What is required for something to be an instance of model-based reasoning is that: 1) it involves the construction or retrieval of a model; 2) inferences are derived through manipulation of the model; and 3) inferences can be specific or generic, that is, they can either apply to the particular model or to the model understood as a model-type, representing a class of models.”

Nersessian describes the mind (the cognitive apparatus) as being capable of “modeling, analogy making, abstraction, visualization, and simulative imagining.” She explains that science has leveraged this and incorporated these approaches into the scientific process (and the scientific method, even). This poses the modeling issue as something that is naturally disposed towards science. This is important and valuable, but also conceals the fact that it is used for other types of reasoning as well, particularly social and creative applications.

Nercessian extensively discusses the history of the theory of mental models, and pays attention to the efforts to equate these models computationally or symbolically. Herbert Simon is specifically mentioned, as an example of someone who applied modeling to computational reasoning systems.

One of the first to develop the theory of models in detail is Craik in the 1940s. The central pillar of Craik’s theory seems to be in mental simulations of models, “reasoning about physical systems via mental simulation of analog representations.” Nersessian mentions that simulation is purportedly developed for navigation within an environment (for instance, a rat in a maze simulates the maze in its mind). Due to human linguistics, mental simulation would thus have the capacity to simulate from language.

Reacting to the idea that thinking is rooted in language, Nersessian suggests that instead, language enables certain narratives, of which individuals may form mental models. Texts describe systems and structures, and these are later manipulated as models and used for reasoning. These are called “discourse models” or “situation models”. This idea seems PERFECT for relating to the narrative adaptation work. This descends from Johnson-Laird, 1989, p471. There is a significant bibliography of sources that discuss how readers form models of texts. The last conclusion reinforces the notion of embodiment and the role of embodiment and perspective within reading a text:

“A number of experiments have been conducted to investigate the hypothesis that in understanding a narrative readers spontaneously construct mental models to represent and reason about the situations depicted by the text (Dijk & Kintsch, 1983; Franklin & Tversky, 1990; Johnson-Laird, 1983; Mani & Johnson-Laird, 1982; McNamara & Sternberg, 1983; Morrow, Bower, & Greenspan, 1989; Perrig & Kintsch, 1985; Zwann, 1999; Zwann & Radvansky, 1998). Although no instructions were given to imagine or picture the situations, when queried about how they had made inferences in response to an experimenter’s questioning, most participants reported that it was by means of “seeing” or “being in the situation” depicted. That is, the reader sees herself as an “observer” of a simulated situation. Whether the view of the situation is “spatial”, i.e., a global perspective, or “perspectival”, i.e., from a specific point of view, is still a point of debate, though recent investigations tend to support the perspectival account, that is, the reference frame of the space appears to be that of the body (Bryant & Tversky, 1999; Glenberg, 1997b; Mainwaring, Tversky, & Schiano, 1996).” (p. 24-25)

I think some degree of this relates to abstract problems or narratives (for instance talking about some blocks of various colors and describing where they are in relation to each other), but it is wholly sensible that these ideas be applied to other conceptual areas, especially in fiction.

Looking at models as they relate to artifacts: People use prosthetics to aid in thinking, demonstrative artifacts that help externalize mental information. One suggestion posed here is that people (specifically scientists in her example) do mental manipulation that interacts with the observed visualization. This process serves to construct a mental model that is *constrained* by the visualization. This process involves a certain coupling between internal and external representations of the model. It also suggests a capacity for the model to be bridged to accommodate other things external to it.

Nersessian goes further to discuss the format of working memory, and propose that the format of the information is 1) modal, and 2) embodied. An example of this derives from 3d positioning, which generally maintains an egocentric coordinate system and perspective. A second example is in the representation of concepts, which Lakoff and Johnson describe as something that is literally part of the brain. The researcher described in detail is Barsalou, who argues that mental representations are perceptual, and man cognitive processes are re-enactments or simulation of perceptual states.