The Emotion Machine (drafts 2003-2004)
by Marvin Minsky (1927-)
This book has not yet been published, or even written. However, Minsky posted early drafts and solicited comments.
The book is something of a sequel to his Society of Mind (SOM), but with a somewhat different aim. SOM was an attempt to describe the mind as a collection of facilities, each of which was in principle implementable in an AI framework; EM is more of a thoughtful look at the nature of the mind, more philosophical, but with an eye to implementation. Unfortunately, implementation is viewed from quite a distance.
The work is (at this time) organized in nine chapters. The last has been released with less than half of its content, judging by its table of contents. Nearly five months elapsed between the release of chapter 8 (and updates to most of the earlier chapters) and the first part of chapter 9.
I sent comments on the first batch of chapters, which seemed to be graciously received. Minsky seemed to appreciate my offer to proofread a later draft. However, it was not until reading chapter 9 that I felt like recording some notes for myself.
The title of chapter 9 is The Self, and he repeats his earlier position that the single-self view, while perhaps useful for ordinary life, is unsuitable for understanding how minds work. Not surprisingly, he believes a network of limited mind-like objects (perhaps I should suggest he call them mindlets), which he calls personae, capable of drawing on memory and formulating questions and answers to other personae, is a model on which to build a workable and working view.
Among the features of his view is the ability to learn Ways To Think, and to control the way we think. Reading this I was struck with the way in which my own thinking has grown disordered over the years, and inspired to attempt to correct this deficiency.
The point of the title is that our minds are largely ruled by emotion. Each major change in the emotional state of our mind as a whole is accompanied by the expression (even if only to one’s ‘self’) of a different sub-personality. For instance, I might be trying to work on my memetic work, and something reminds me of an emotionally painful episode from my past. The switch to attending to the memory is accompanied by changes in attitudes, even body attitude, related to the different state of mind in the two different kinds of thinking.
Minsky wonders about traits: why is there such a thing as personality traits? To start, consider a list he uses: disciplined, honest, attentive, and friendly. Society reinforces the recognition of these traits. If we value them in others, we also might try to develop them in ourselves. If I wish to be disciplined, to acquire the Ways To Think associated with being disciplined, I will examine that trait and find activities associated with it, such as keeping prioritized lists of things to do; the activities might include meta-activities, such as reviewing and updating the lists of things to do.
In fact this is just what I have decided to try. In the past I have dismissed the utility of such schemes as the Franklin Planner, with its orderly lists. However, the principles behind it are clearly conducive to developing disciplined Ways To Think. It seems worth a try.
I will create a new report with more material related to the content of the book after it comes nearer completion. At that time, I might also comment on the results of my endeavor to become more disciplined in my thinking.
Note that I also started reading the finished book in 2007: