Thursday, 28 May 2009

Mental Model Literature Review - Part 1

In this post I not only wish to contextually review the literature on mental models but also to define the use of the term that will be central to my first practical PhD project. Mental Models is unfortunately a term that has had its meaning diluted by non-cognitive science practioners such as Indi Young (more on this later). Therefore I will be exploring Mental Models firstly from a cognitive science perspective in order then to draw parallels and influences into Interaction Design. This post is not a defined thesis on my part but an attempt to draw together the essence of mental models. There will be theoretical holes in what I will post, so feel free to point them out. Look on what follows as simply my first attempt to stitch together my understanding of a very complex cognitive science discourse, from a discipline that is not my own. I will continue this literature review across several posts with the aim of refining my understanding.


To begin with I am going to quote psychologist Kenneth Craik (1914-1945) from his 1943 book The Nature of Explanation. Craik is cited by cognitive scientist Johnson-Laird in his 1983 seminal work Mental Models as proposing the first modern hypothesis on humans being processors of information in order to understand phenomena.

”If the organism carries a “small-scale model” of external reality and of its own possible actions within its head, it is able to try out various alternatives, conclude which is the best of them, react to future situations before they arise, utilise knowledge of past events in dealing with present and future, and in every way to react in a much fuller, safer, and more competent manner to emergencies which face it.” (Craik, 1943, p57)

This statement was to open up a whole new area of research within the already new scientific discipline of cognitive science. By 1983 Mental Models research had two seminal works that explored this concept. Johnson-Laird proposed that the human mind uses propositional reasoning instead of reductive mathematical logic to understand phenomena. Gentner and Stevens, in their edited collection of papers, also entitled Mental Models, present a diverse range of views, including an early paper by Dr Donald Norman who here begins the discourse upon the Design Model. If Johnson-Laird proposes propositional reasoning, then Gentner and Stevens' book explores the structure of knowledge representations specific to different domains of application. Johnson-Laird states in introducing his book,

"The psychological core of understanding, I shall assume, consists in your having a 'working model' of the phenomenon in your mind. (…) you have a mental representation that serves as a model of an entity (…) A model has, in Craik’s phrase, a similar ‘relation-structure’ to the process it models, and hence it can be useful explanatorily; a simulation merely mimics the phenomenon without relying on a similar underlying relation-structure. Many of the models in people’s minds are little more than high-grade simulations, but they are none the less useful provided that the picture is accurate; all representations of physical phenomena necessarily contain an element of simulation." (Johnson-Laird, 1983, p2/4)

The importance here is that a 'relation-structure' the functionality of a cognitive conceptual model. To propose an explanation on how a phenomenom works, a model that is constructed by a person is NOT improved by them also trying to understand the specific context it is encountered in. The specific context is not transferable to another context but the mental model will be. The model will not be complete each time it is applied, but in Chomsky's term 'explanatory adequate'. To Johnson-Laird because a mental model is constructed by mental logic, mental logic has six main problems. Inference, formulation and deduction all are contributory to the empirical nature of mental logic. He lists the problems as (pp39-40):

  1. People make fallacious inferences.

  2. Which logic, or logics, are to be found in the mind?

  3. How is logic formulated in the mind?

  4. How does a system of logic arise in the mind?

  5. What evidence there is about the psychology of reasoning suggests that deductions are not immune to the content of the premises.

  6. People follow extra-logical heuristics when they make spontaneous inferences.

People's powers of deriving a reasoned solution based upon a premise they have, or from precedence is strongly affected by their own cognition that is deceptive. Mental models, although helpful to understanding a phenomenon, are inherently 'incomplete' (Johnson-Laird, 1983)(Norman, 1983), 'doubtful validity' (Norman, 1983), 'fragmentary' (Norman, 1998), 'dynamically constructed' (Preece et al, 1995), and 'not always correct' (Gentner and Gentner, 1983). As Johnson-Laird argues, the formulation of a mental model to understand a phenomenon is 'propositional reasoning' on behalf of each individual. Norman explains this in more accessible terms:

"The power of mental models is that they let you figure out what would happen in novel situations. Or, if you are actually doing the task and there is a problem, they let you figure out what is happening. If the model is wrong you will be wrong too." (Norman, 1998, p71)

If the person's proposition regarding the causality, relationships and mechanisms within a phenomenon encountered is poorly appreciated then their understanding of that phenomenon will not work. But an incomplete proposition that does capture the behaviour of that phenomenon, that for them represents the causality and mechanisms of that phenomenon based upon previous experiences, or from fragmentary experiences of aspects of similar phenomena, can arrive at an 'explanatory adequate' mental model. This is why Norman describes the human mental process as "unscientific" that supports Johnson-Laird's argument that humans do not use reductive mathematical logic as the brain is a computer. In their book The Embodied Mind Varelo et al refer to phenomenologist Edmund Husserl (1859-1938) to contextualise human cognition from a different perspective, where a human's 'sensorimotor capacities' is enabling enough to successfully interact within an environment with the phenomenon. According to their thesis Husserl claimed

"that to understand cognition, we cannot take the world naively but must see it instead as having the mark of our own structure. He also took the second step, at least partially, in realizing that that structure (the first step) was something that he was cognizing with his own mind.(…)He explicitly focused on the experience of consciousness in what he called the “lived-world.” The lived world is not the naive, theoretical conception of the world found in the natural attitude. It is, rather, the everyday social world, in which theory is always directed toward some practical end. Husserl argued that all reflection, all theoretical activity, including science, presupposes the life-world as a background. The task of the phenomenologist now became the analysis of the essential relation between consciousness, experience, and this life-world.” (Varelo et al, 1996, pp16-17)

For each person to construct a mental model of a phenomenon based upon proposition reasoning, entails them cognizing the problem from their previous experiences built into a appropriate functional model (Preece et al, 1995) developed from existing knowledge of a similar context. As a mental model is a dynamic conceptual tool to solve a problem within a phenomena, they can be be said to be drawn from Husserl's concept of the 'lived-world'.



See Design Model - statement for list of references cited in this post.

No comments: