I've never had any education in psychology/sociology, except for some books I read on memetics and anthropology, but I'd like to share a model I've developed of human psychology/sociology since I've gotten interested in the subject. I realize that this is a game development forum, but programmers are the best people to discuss conceptually-demanding ideas and functionality-oriented thinking. Or at least some might be interested in applying some concepts I discuss here to games. At the core of the human mind are drives. These are the deep-rooted animal instincts that we are all born with. Some examples: - hunger - sleep - social interaction - social power - sex - curiosity - etc... Satisfying these drives isn't as simple as "go up to the tree and get a banana". Homo sapiens would've never survived the cold winters if they weren't capable of quickly adapting to changing environment and following a complex strategy to attaining their needs. This general strategy changes and grows over a lifetime as the human animal has new experiences. The general strategy consists of super-goals, which are broken down into goals, which are broken down into sub-goals, and so on. The actions flow down the chain of intent, and the rewards/emotions flow back up the same path to validate/reinforce or invalidate/discourage the behaviour/strategy. An example of a strategy: > am hungry -> need to eat --> need to buy food ---> need to get money ----> need to get a job -----> need to work hard The general strategy is the behaviour of the human. It can be thought of as the software that runs the human hardware. It is hotwired into lower-level systems (conditioned responses, instincts) by a functionally-complex network. The human operating system has the ability to rewrite its programs, but not the hardware. It can change what it eats, how it chews, but not what foods taste good. Naturally, a certain amount of drives must be satisfied in day-to-day life to maintain sanity, but the human animal is mostly free to experiment and pursue a long-term general strategy for attaining happiness. The general strategy can be broken down into smaller parts that deal with attaining specific goals. Different strategies are competing for use by the person. For a given intent, different strategies can be employed at different times under changing circumstances. But it's not a top-to-bottom pyramid (functionally speaking). It resembles a giant mesh, with feedback loops, links back up, small clusters of tightly-knit units, etc. Ideally, the perfect strategy satisfies the most instincts the best for the least effort. Person naturally follows the strategy that's most successful. For a given intenet, the sucess of a possible strategy can be thought of as an equation: success = rewards / effort It can be thought of like running a business. Goal is to get money (happiness). Money (happiness) flows in as a result of doing work (actions), which required money (effort) to be spent in the first place. Happiness/energy/will-power is never actually spent/destroyed, but converted to internal drives/emotions/instincts. I suppose emotions are conserved, same way as energy or matter. Happiness/effort/will-power begets more happiness/effort/will-power. When a person is happy, he has more energy to experiment with other strategies and be creative. When rewards are diminishing (more effort is required, but less rewards are gotten), then person becomes sad, tired, depressed, etc. When a person is low on happiness/energy/will-power, he conserves internal resources, avoids confrontations, avoids unnecessary expenditure of effort, and chooses the strategy that takes least effort. This creates the basis for stable social classes. Those on the bottom have less happiness/will-power/energy and so must invest it on satisfying more essential drives, which overpower less essential drives. Those on top have an abundance of happiness/will-power/energy and can afford to pursue more effort-demanding actions, like consolidating social power and engaing in confrontations. Happiness also functions as a social signal. It tells others "I can satisfy my drives with little expenditure of effort". This makes others envious/respectful/friendly to this person (depending on social relation), and makes them try to imitate the same successful strategies/behaviour for themselves. Faking emotions goes against this social mechanism by distorting the success rate of strategies/behaviour. Other people percieve intent of the subject carrying out an action, and internally calculate the expected success rate. If expected success rate differs from the one transmitted emotionally by the subject, then the subject is "faking it" for the purpose of some other, social intent, and it is punished instinctively. Sadness, gloominess, depression, etc. also function as social signals. They tell others "my strategy cannot satisfy my drives without using up lots of effort". Emotions also function as social unifiers. A social group (short-term grouping) is united by emotions. They conduct the same emotions, like electric currents. Everyone is emotionally dependant on everyone else. This reduces internal disorder within the group so they can compete as a macro-organism to achieve the same intent/goal (like, bringing down an animal, killing the other tribe and stealing their women, etc). There are those who radiate/transmit the emotions (usually those most respected), and those who internalize/conduct them in a group (usually those least experienced). The group members form an emotional circuit. Imitation and communication facilitates the transfer of strategies and the information on which the strategies rely. That's enough for now... sorry, I realize the wording is wierd, but I'm finding it hard to express the ideas without destroying the meaning. So, does this model reflect/parallel how the current psychological theories picture the human mind? Do you see any theoretical flaws in my model? Any logical conclusions to derive from my model? Anything to add to the model?

I don't believe that our actions are purely controlled by 'drives' as you state. One of the most prevalent psychological models attempting to explain similar behavior is Maslow's Pyramid, which even says that drives are only integral at the lower stages of the pyramid.

Two interesting bits of evidence against purely the drive theory are with the hunger and sex drives. Note that as the human race developed, we have moreorless gotten rid of the hunger instinct (amongst most classes in the US at least). We eat three meals each day systematically, and few of us have actually experienced this drive. Second, the sex drive is said to be the only 'optional' drive -- if we do not satisfy the lusts, then there appears to be no ill effects on life. In both counts other factors other than the drives are acting.

The model I use is that the brain has many inputs. We have visual, tactile, auditory, olfactory, and gustatory external inputs. However, we also have internal feedback inputs, "drives" -- hunger, thirst, sex, sleep, etc. I argue though that the definition of a drive is not as liberal as you mention, but rather something that you can feel yet cannot control without external interactions. Drives are just another input into your body, and you act on them similarly to any other input.

Toward this, think about pain. This is not a 'drive' (it is clearly physiologically stimulated), but it has a great effect on your actions. It is an input into your body which you give preference to because it is an uncomfortable feeling. Similarly, you give preference to many internally-generated signals because they provide uncomfortable feelings.

Your model is interesting, but very vague. AI folks have been working on purely splitting tasks into sub-goals to little avail. Further, yes, drives have an influence on 'general strategy' and on 'actions', but it is in determining exactly how all of those wires are connected where the mystery of the brain lies. If you can produce a computational model which can make predictions, then I think it would get a lot more attention.

I'm working on something like this right now, actually. We're using Maslow's Hierarchy as a starting point, but allowing for the exceptions that would violate it (such as the "starving artist"), and we're including Keirsey style (see also Myers-Briggs) temperment type in determining actions. Basically, an individual has his own -way- of deciding how to fullfill a need.

We're still mostly at the planning stages, only Engine code done so far, but I can tell you it's probably not going to be fast - lots of calculations involving lots of trees.

If I were you, I'd do a lot of reading on Maslow, and also on the personality theory (Keirsey and Myers-Briggs). For me, the Maslow stuff helped me organize my need tree quite a bit and simplify some of the ideas into different layers.

Quote: Original post by Rhalin
I'm working on something like this right now, actually.

This sounds interesting! Is it as a graduate project (and in what major if so -- Mathematical Psychology?) or a hobby adventure?

Quote: This sounds interesting!

Indeed. I'd love to hear more about this project.

Maslow's model looks like a good starting point to learn about psychology. Thanks for the info.