Having taken the trouble to provide him with answers I thought I might as well put them here, in case anyone is interested.
[Occasionally I'll add or modify this page, using square brackets to indicate such modifications.]
I may well offend some people, especially if they have fallen victim to the fallacies and wishful thinking that I claim afflicts many people who attempt to investigate the role of emotions in intelligence, as explained in this slide presentation: DO MACHINES, NATURAL OR ARTIFICIAL, REALLY NEED EMOTIONS? (originally prepared for a lay audience at a 'Cafe Scientifique' meeting).
[Added: 29 Sep 2006
My answers to the questions are all based on the use of the design-based approach to the study of mind. ]
The most general concept that fits our fuzzy and indeterminate mish-mash of uses of the words 'emotion' and 'emotional' is:Example: a house-fly consuming food detects something rapidly descending towards it, and the 'alarm' mechanism aborts eating and triggers escape behaviour. In humans there is a far wider variety of cases involving evolutionarily older and newer mechanisms. E.g. a mathematician working on an important new proof notices the possibility of a fallacy caused by implicit division by zero. This may trigger a disposition to switch to investigating the offending step in the proof. Some of these disruptions can be unconscious -- like the people who are jealous or infatuated and don't realise it, though it is evident to their friends.A state in which a monitoring mechanism acquires a tendency (i.e. a disposition, possibly suppressed) to abort, redirect, or modulate some other process or collection of processes.
In the obvious cases the tendency is not resisted and some change occurs as a result. In more subtle cases the disruptive tendency may be suppressed or overridden, but it is still there competing for control.
[ In 1996, Ian Wright, Luc Beaudoin and I published an analysis of long term grief, which is one of the most common examples of a state commonly labelled as an emotion which can endure over time, even while completely different emotions occur, e.g. enjoying a joke, concern about losing one's job, falling in love, etc. Other examples of states that can endure while temporarily suppressed are jealousy, infatuation, anger, concern about one's government's actions, intense support for a political movement, joy at having a new baby, excitement about a research project, and, on a shorter time scale, excited anticipation of a forthcoming event.(I personally think that 'emotion' is not a concept that has sufficient precision/clarity/uniformity of usage, to be useful for scientists.Theories of emotion that do not allow for the possibility of such phenomena must be false, unless their proponents hi-jack the label 'emotion' by re-defining it to suit their theories. ]
The main function of the mechanisms referred to above is to prevent 'normal' processing from continuing in circumstances where some state requiring (or prima-facie requiring) a change of 'direction' occurs.[Other things that are labelled as 'emotions' by other people, e.g. desires, concerns, attitudes, preferences, moods, pleasures, pains, etc. have other functions.]
Concurrently active mechanisms, namely[Some of the monitoring mechanisms have to be fast in reacting to dangers and opportunities, so they will typically be fairly shallow, and they can produce both false positives and false negatives. Bad 'training' can lead to highly dysfunctional processes.](a) those doing some task,
and
(b) those monitoring performance of (a) and the environment, using fast, trainable pattern-recognition mechanisms.
Not sure what you mean. In my work I have talked about variable-threshold attention filters that suppress interruptions and disturbances, at least in part of the system. The filter threshold will depend on urgency, importance, and resource requirements of current tasks.That's an answer regarding regulation of initiation.
Once emotional disruption has been initiated there are other kinds of regulation required, e.g. not over-reacting to accidental behaviour in a young child.
There's also long term regulation which is part of learning -- e.g. training the monitoring mechanisms.
For a deep science of mind, covering humans, other animals, and robots, what sort of ontology of affective states and processes is needed?[The ontology will be different for different sorts of animals and machines, and for humans at different stages of development, or with different kinds of brain damage or disease. A meta-ontology will show how different architectures support different mental ontologies.]
Yes(Also some slow, deliberate, and conscious connections, e.g. when people learn to detect and control their emotional reactions -- which does not happen often enough I fear.)
Last updated: 26 Apr 2008