Talk:CCNBook/Main

From Computational Cognitive Neuroscience Wiki
Jump to: navigation, search

Getting The Discussion Started ...

I would like to contribute to the CCN Book revision. I have created an account on this Wiki. Is there anything else that I should do in preparation for making use of this online tool? Is this the right place to be having this discussion?

In previous conversations, it was suggested that this book would be more attractive if real cognitive phenomena were addressed earlier, rather than using mostly toy problems to illustrate basic concepts and pushing all of the "real" models off to the second half of the book. At a glance, it looks like this suggestion has been vetoed. The old structure of the book is still present. Perhaps it would be good to discuss the desired high-level structural issues first, making it easier for individual contributors to figure out where their contributions would fit.

--David C. Noelle, Ph.D. 01:51, 19 January 2011 (MST)


Definitely not vetoed. I think it is a bit tough to figure out how to do that successfully though: lots of cart before the horse issues. In my intro lectures this time around, I gave a quick tour of lots of the cognitive phenomena, and tried to give a big picture framework of where we're going. But at some point you need to talk about what the neuron does, and that is difficult to do in an large-scale model.. So, my current sense of things is the following:

  • each of the "part I" fundamentals chapters should be much shorter, and really focused on getting the key points across with a minimum of fuss and bother. There will be lots of "extra" content for those who are interested, but it should be possible to cover this material much more quickly than in the past.
    • move all the TD/RL stuff to a separate chapter in part II that covers this stuff in full detail with relevant bg/vta etc biology.
  • then there is more time for the cognitive stuff, and it comes more quickly. and lots of the relevant content e.g. from the "networks" chapter can be covered in the later domain-specific chapters, where relevant. networks in particular is due for a massive slim-down.
  • the later cognitive models will be presented in a fairly "standalone" way that makes it easy for people to just assign individual models, even while skipping much of the fundamentals stuff, so that individual instructors could decide what to focus on and what to skip -- could teach a pure "phenomenology" class without any fundamentals. also, people teaching other classes entirely can just grab models of interest as purely ala-carte pieces. (and then get hooked and come back for more!).
    • we can have a very brief "summary of fundamentals" section that all the cognitive models can point to, which provides a super-succinct summary of the mechanisms, with pointers to the more thorough sections.

Bottom line idea is to make everything more modular and configurable for individual instructor's preferences. I think the wiki technology enables this kind of thing really nicely, and we should leverage it! For example, people can design their own ordering of everything on a "book view" wiki page, that represents the way they want to approach the info. Others can shop around for these different views and use the one's they want, or create new ones.

Last point here is that the current goal in this very first draft is just to get through the semester, realistically speaking, and create a first-pass vision of material that fits some of the above goals, but can then be massaged in various ways after that to achieve the broader goals that emerge as we continue these kinds of discussions.. Again, the wiki should enable things to be easily reconfigured in various ways as we go through. There just isn't time to redo the simulations and the text in this iteration, so that places a big constraint on things: need to make this version work with existing sims. Next iteration will redo the sims and update text around that.

So, bottom line, probably best contribution at this point is to make suggestions at all levels and come up with new ideas for what and how to present, and edit content as it comes online, and then we can take stock after the semester is over and have more of a master plan for where to go from there. This is pretty seat-of-the-pants but I'm not sure how else to do it without spending a lot of time that is not readily avail in the planning stages..

--Oreilly 02:25, 19 January 2011 (MST)