On the Origin of Concepts - Part 2

In the first part of this extended post, I discussed how I've found biological research patterns to be a useful analogy when doing my work in cognitive education research. I started by trying to characterize the concept of biological evolution, working on a hunch that it was a coordination class concept, which itself was based on a hunch that coordination class concepts can demonstrate the four classical patterns of conceptual change. Although I quickly found that biological evolution fit with many of the criteria for the coordination class model, I got in my own way and started looking at the model pretty closely. I couldn't shake a couple of major thoughts, both of which were rooted in my prior education experiences in neuroscience, developmental biology, and teaching. First off, the coordination class model looks a lot like neural networks, which is a good thing, since - so far - we're making claims about human thinking and learning, and we know from neuroscience that those functions are generated by the brain. Second, I've not yet discovered a conceptual model with an internal and inherent mechanism for learning - that is to say, I kept wondering how it was that a coordination class concept actually changed in learning experiences. Third, I've witnessed a lot of different learning experiences for a number of different high school students first learning about the concept of biological evolution, and I feel totally confident in saying that personal beliefs about the nature of knowledge (fancy: personal epistemology) are part of the conceptual structure and strongly influence the way in which a concept develops - especially the concept of biological evolution. I've also been a part of some incredibly creative learning experiences, and it seemed to me that students with the strongest inventive thinking skills were also the students that formed the most expert-like concepts.

To start answering my first observation that coordination class concept graphs look a lot like neural networks, I did a bit of research into neural networks and learning. Needless to say there is a lot of research on this topic in vivo, and there are also many efforts to model neural networks as abstractions in computer software. One of the interesting ideas that turned up in this process was the idea of scale-free networks. These are, in essence, networks of objects that retain certain properties at different levels of size and complexity. Interestingly, these types of networks exist at (at least) the macromolecular, cellular, and cognitive levels of complexity.

To start answering my second observation that coordination class concepts lacked an internal mechanism for learning, I began researching conceptual change theory literature. I found a simple theory on conceptual change that relied on three progressive processes: plausibility, comprehension, and fruitfulness. Based on my teaching experience, I chose to use these processes as functional categories for different types of parts of thinking in science: plausibility is linked to inventive resources (like imagination), comprehension is linked to sense-making resources (like processing time scales), and fruitfulness is linked to epistemological resources (like knowledge-as-transmitted-stuff). Further literature review has demonstrated that each of these categories can be divided into smaller categories. So, I've started to think of this as the development of a taxonomy for the parts of thinking.

More to come - time to get ready to teach.

On the Origin of Concepts

So, a great meeting with Michael today. I knew we'd hit it in stride right from the get-go when he started off our conversation with an update on cognitive resources. One of the major topics of discussion at the first Cognitive Group meeting this semester was the nature of the components of thinking (see my "farewell to p-prims" post). Word from Michael's contancts in the University of Maryland group is that the term "resource" was created with the intent of allowing for larger scale representations of thinking (so, a concept can be a resource for a larger concept), and that they are not necessarily primitive (though they can be). So, I think that pretty much settles it for me: p-prim, while a useful functional descriptor, isn't going to be useful enough for my purposes. Up until this point, I think I've been fairly vague on that - purpose, intent - but I think today was one of those great times in academic life when you really, really refine your work.

So here it is: I'm developing a taxonomy for cognitive resources, and in so doing, I'm also creating a method for determining and evaluating conceptual complexity. I'm doing this because I am interested in developing a model for the concept of biological evolution that is based on a knowledge-in-pieces view of cognition, and I've been frustrated by the formal, functional, and developmental limitations that are characteristic of existing models for concepts and hypotheses on conceptual change. Because of my frustrations in the modeling of biological evolution from a cognitive perspective, I've applied classical biological methods in the cognitive realm: anatomical dissection (identifying the pieces of thinking and their connections), physiological mechanism determination (indentifying the functions and functional mechanisms of pieces of thinking), developmental classification (characterizing changes in thinking over time, the mechanisms for change, and the influences on change), and taxonomic organization (classifying the formal and functional similarities and differences between and among multiple aspects of cognition).

Next - why this is going to be very useful.

Epistemological Resources and a Developmental Assessment

Steelers 21, Seahawks 10. I was rooting for Seattle since it was their first Super Bowl appearance, but it's good to see the Lombardi Trophy stay in the AFC. Maybe next year will bring some better post-season play from the Pats.

Anyway...

There were a total of four papers that I found the other night at UM, and the next one I want to talk about is by Tsai and Liu, called "Developing a Multi-dimensional Instrument for Assessing Students' Epistemological Views toward Science" and published in the October 2005 issue (v27, n13) of International Journal for Science Education. In this paper, students used Likert scales to "agree" or "disagree" with a number of statements that probed their epistemological view of science. The "multi-dimensional" aspect is really just the further differentiation of epistemological resources into five major categories: social negotiation, inventive and creative nature, theory-laden explanation, cultural impacts, and changing and tentative features. Within each of these categories are a minimum of three ideas (I'll include them in my model as specific resources); social negotiation has 6, inventive and creative nature has 4, and the other three categories each have 3 resources within that category.

Imagination as a Cognitive Resource

I recently stumbled upon a few references to the role of imagination in science learning. The first "hit" turned up in a review of recent editions of the International Journal of Science Education. In the April 2005 issue (Volume 27, Number 5) I found an article by James H. Mathewson called "The visual core of science: definition and applications to education" (pages 529 - 548). [note - I can't find any home page for James H. Mathewson on the SDSU site or elsewhere - he's doing interesting work, so I'm hoping to turn up more info in the future. another note - SDSU apparently has a program that is very similar to UMaine's MST - it's called the Center for Research in Mathematics and Science Education, and features a Ph.D. option.] In the past, he's also written an article called "Visual-spatial thinking: an aspect of science overlooked by educatiors", which appeared in Science Education, v83 n1 p33-54 Jan 1999. I have not yet read this article. Also of note, I was looking through the references in "The Visual Core of Science" and found a book that turned out to be very interesting. The author is Arthur I. Miller, and the book is titled Insights of Genius: Imagery and Creativity in Science and Art, published by MIT Press (paperback in 2000).


So, a brief mention of some of the aspects of the Matheson paper that were most interesting. First is the notion (developed by Gerald Holton - not yet researched) that imagination has three different functional sub-categories: visual, metaphoric, and thematic. I am going to map these as resource categories within the imagination category.


OK, well, Super Bowl parties are 'bout to start, so I'll be catching up on this post tomorrow.