Tutorials

Tutorials allow conference attendees to expand their knowledge and can introduce researchers to emerging topics, new technologies, or present an overview of the state-of-the-art in a particular area

Diagrams 2014 will host one tutorial (the presenter of the second is no longer available):


Semantic Properties of Diagrams and Their Cognitive Potentials

Instructor:
Prof. Atsushi Shimojima
Faculty of Culture and Information Science
Doshisha University, Japan

Tutorial length:
1 hour

Tutorial background:
Content In modern discussions of how diagrams facilitate, fail to facilitate, or misguide our thinking and understanding, several themes have appeared sporadically but quite recurrently. Each of these themes appears as a more or less formal psychological description of a cognitive function of diagrammatic notations, but as it turns out, they are all rooted in logically characterizable semantic properties of the relevant diagrammatic notations. The primary purpose of this tutorial is to demonstrate this finding, by referring to numerous examples with which the efficacy and inefficacy of diagrams are discussed in the literature. We then classify them into a small number of general cognitive phenomena, and dig out the semantic property of graphical notations responsible for each of them.

The tutorial is an refined and expanded version of the tutorial of the same goal given in Diagrams 2004.

The tutorial will serve as a fairly comprehensive survey of the literature on the cognitive functions of diagrammatic representations. Such a survey/tutorial is especially important since the works in this area are typically scattered over diverse fields such as AI, cognitive psychology, philosophy, logic, and information design, conducted in different methods, vocabularies, and degrees of technicality. This prohibits an easy overview of various results, proposals, and suggestions offered in the area. The audience will obtain an accessible summary of these results and ideas, described in a single, systematic conceptual set.

Tutorial topics/themes:

  1. Free ride properties:
    Expressing a certain set of information in the system always results in the expression of another, consequential piece of information. This theme has been suggested or proposed, under various names, as an explanation of certain automaticity of inference conducted with the help of diagrams (e.g., Lindsay 1988; Sloman 1971; Barwise and Etchemendy 1990; Larkin and Simon 1987, Shimojima 1996).
  2. Auto-consistency:
    Incapacity of the system to express a certain range of inconsistent sets of information. The theme has been suggested as an explanation of the ease of consistency inferences based on diagrams (e.g., Barwise and Etchemendy 1994; Barwise and Etchemendy 1995; Stenning and Inder 1995; Gelernter 1959; Lindsay 1988; Shimojima 2002).
  3. Over-specificity:
    Incapacity of the system to express certain sets of information without choosing to express another, non-consequential piece of information. The concept has been suggested or proposed as an explanation of the difficulty of expressing "abstract" information in certain diagrammatic systems. (e.g., Berkeley 1710; Dennett 1969; Pylyshyn 1973; Sloman 1971; Stenning and Oberlander 1995; Shimojima 1996).
  4. Meaning derivation properties:
    Capacity to express semantic contents not directly prescribed by the basic semantic conventions, but only derivable from them. The concept has been offered as an explanation of the richness of semantic contents of graphics in certain systems. (e.g., Bertin 1973; Kosslyn 1988; Tufte 1990; Pinker 1990; Shimojima 1999; Ratwani et al. 2008).

Depending on time available, we will extend out discussion to such themes as envisaging (e.g., Hegarty 1992; Bauer and Johnson-Laird 1998; Schwartz 1995; Shimojima and Katagiri 2013), aspect shifting (Jamnik 2001; Giaquinto 2007), and law-encoding diagrams (e.g., Cheng 1996, 1997, 2002).

Intended audience:
Graduate students and professional researchers in the following categories: philosophers interested in semiotic characteristics of diagrammatic notations, logicians interested in formalizing inferences based on diagrammatic notations, AI researchers interested in realizing diagrammatic reasoning capacities in their systems, cognitive psychologists interested in cognitive processes involved in the use of diagrams for comprehension and reasoning, computer scientists interested in visualization of computational processes ("visual languages"), and theoretically oriented graphic designers interested in cognitive potentials of diagrammatic notations.

Instructor background:
The instructor received his doctoral degree from the Philosophy Department at Indiana University in 1996. He was researcher at Advanced Telecommunication Labs in Japan, associate professor at Japan Advanced Institute of Science and Technology. He is currently professor of the Faculty of Culture and Information Science at Doshisha University. He is also research associate of the Openproof project at the Center for Study of Language and Information at Stanford University. His research has been focused on applying semantical analysis of diagrammatic systems to make predictions on their cognitive potentials and verifying these predictions on the basis of psychological experimentation.

The following is the list of the instructor's publications closely related to the content of the proposed tutorial:

  • "Operational Constraints in Diagrammatic Reasoning." G. Allwein and J. Barwise Eds. Logical Reasoning with Diagrams. Oxford: Oxford University Press, pp. 27–48. 1996.
  • "Derivative Meaning in Graphical Representations." The Proceedings of 1999 IEEE Symposium on Visual Languages, pp. 212–219, 1999.
  • "Constraint-Preserving Representations." L. S. Moss, J. Ginzburg, and M. de Rijke Eds. Logic, Language and Computation, Volume 2. Stanford: CSLI Publications, pp. 296–317. 1999.
  • "The Linguistic-Graphic Distinction: Exploring Alternatives," Artificial Intelligence Review, Vol. 13, No. 4, pp. 313–335, 1999.
  • "A Logical Analysis of Graphical Consistency Proofs." L. Magnani and N.J. Neressian, Eds. Logical and Computational Aspects of Model-Based Reasoning. Dordrecht: Kluwer Academic Press, pp. 93–115. 2002.
  • "The Inferential-Expressive Trade-Off: A Case Study of Tabular Representations." Mary Hegarty, Bernd Meyer, and N. Hari Narayanan Eds. Diagrammatic Representation and Inference: Second International Conference, Diagrams 2002, Proceedings. LNAI 2317, Springer, Berlin, 116–130, 2002.
  • "What Makes a System Less Graphical?" Machine Graphics and Vision, Vol. 12, No. 1, pp. 99–115, 2003.
  • (With Y. Katagiri) "An Eye-Tracking Study of Exploitations of Spatial Constraints in Diagrammatic Reasoning," Cognitive Science, Vol. 37. No. 2. pp. 211–254, 2013.
  • Semantic Properties of Diagrams and Their Cognitive Potentials, Book in preparation.