A Foundation of Ethics

Rob Kamp, Netherlands

Version 1.1.0: July 2007 (Version 1.0.3: May, 2005)

The theory is a set of definitions, principles and postulates and a few examples.

Ethics and philosophy

It seems to me that current philosophy consists of the following subjects. The subjects are:

• Ontology. The study of taxonomies of properties, relations and processes (attributes). (UML could be used.)
• Epistemology. The study of (the justification of) knowledge (true believes and the metacognition that the believes are true).
• Ethics. The study of behaviour according to the ethical principle. (See below).
• Logic. The study of truth relations between propositions.
• Foundations of sciences. The studies of basic definitions, principles and postulates of a certain science.
• Sociology of science. The study of how scientists influence each others goals, plans and behaviour.
• Anthropology of science. The study of customs (common habits) of a (subgroup of a) scientific community.

A definition of ethical behaviour

It seems to me that the most objective, intuitive and practical definition of ethical behaviour is a consequentialistic one. First some background definitions.

• The strict definition of behaviour is a perceptual change of the body of a person. Examples: change of posture; change of facial expression; movement of arms and legs; breathing; speeking; sweating; blushing.
• According to the strict definition facial expression is not a behaviour. It's a state of the face (muscles). Change of facial expression is a perceptual process of the body and thus a behaviour. Strictly speaking a static facial expression is not a behaviour. But a more practical definition of "behaviour" is an extended one. It includes the perceptual static states of the body of a person. Thus according to the extended definition, not moving at all, is also behaviour.
  A behaviour can influence the mental or physical state of a person. Directly or indirectly: via its environment or by his own intervention.
• Environment of a person is defined as the matter-with-its-structure in any space that encloses the person. Somebody may be in his car driving home, but he can also be at home. That's his environment.
• Event is defined as a temporary attribute in an environment. An event has a (relatively short) duration.
• Circumstances of a person is defined as some events of his environment and/or of the person that are not characteristic of him. Somebody's height or hair colour characterizes him, but whether he has catched a cold or is sleepy, and that it's late are his circumstances.
• Situation is defined as a temporal sequence or flow of circumstances of a person. A situation has a duration.
• Actor and receptor. An actor is a person that acts. A receptor is somebody whose circumstances or personal characteristics are affected by the action of the actor. Other terms that may be useful for analyzing a situation are (toolbox): transceptor (mediator if animal); source; transporter; conductor; transistor; transformer; transactor.
• Rationality of an actor comprises: the number of steps it plans ahead; its morality principle (see below); what he knows of its receptors and/or actors and expected happinesses (H_e as a result of his possible actions. An actor A wants to make a prediction of how another actor B responds. He wants to know the rationality of the other actor B. In the social science of ethics it can be statistically researched how rationality, personal characteristics and situations correlate. This data can be used to make predictions of how certain persons react in certain situations (social psychology).

The aim of ethical theory is to be a guide for individual decision making given the consequences for the receptors (which may include the actor himself). The consequences may be caused with or without intervention of the receptors themselves (in certain degree). There are in each step certain alternative behaviours from which an actor can choose. (The reaction of the receptors is given.) The (most) ethical alternative is the one that satisfies the ethical principle. From the ethical principle there may be derived behavioural heuristics to react in situations (special cases).

The ethical principle is (loosely speaking): Minimize unhappiness in number and in degree as much as possible. Every creature with feelings counts. Animals (including man). A more precize defintion: During or after each alternative somebody (or a group of people but also animals) has a minimum degree of happiness. The most ethical behavriour is the one such that its minumum degree of happiness is a maximum relative the other minima. (Not each happiness needs to be caused by the alternative action).

Example:

• Alternative behaviour A: Result: Happiness(Pete) = -1, Happiness(Jan) = +1, Happiness(Jim) = 0.
• Alternative behaviour B: Result: Happiness(Pete) = 0, Happiness(Jan) = -2, Happiness(Jim) = +3.
• Alternative behaviour C: Result: Happiness(Pete) = -3, Happiness(Jan) = +4, Happiness(Jim) = +5.

Another example:
Short notation: A = (-1,+1,-1), B = (0,-2,+3), C = (-3,+4,+5). Minimum(A) = -2 (!), Minimum(B) = -2, Minimum(C) = -3. In this case there is no maximum. Behaviour A and B are equally ethical. Note that the minimum of A is twice -1. M(A) has a exclamation mark because superficially one would expect M(A) to be -1. But in this case the group of people with least H (namely -1) comprises TWO persons. This makes the TOTAL H of the group: M(A) = 2 x -1 = -2.

Some more definitions:

• The Feeling/Pleasantness quantity (F) is the degree of overall (un)-pleasantness a person experiences. It has a continues scale between -10 to +10. The scale is a ratio scale. It is symmetric with respect to intensity (quantity) and anti-symmetric with respect to pleasantness (quality). An F-value corresponds to two complementary feelings: moral emotions and other feelings (including non-moral emotions). A moral judgement is a (dis-) approvement of a sitiuation or an event (especially a behaviour). In case an actor disapproves of something he may have negative feelings: dissatisfaction, he may be angry, sorrow. If he approves of a behaviour he may have the moral emotions of being content or being happy. Only when an actor deliberates a (possible) behaviour he may have moral emotions. If he reacts to a behaviour with an emotion (e.g. feeling insulted) or if he is actualizing a behaviour he may have at that moment no moral judgements and thus no moral emotions. The degree of pleasantness of a moral emotion is a moral attitude. Thus an actors moral attitude contributes to his F-value. Attitude is the degree of (un-)pleasantness (F-value) of feelings (including emotions) that are the result of judging the desirability (P-value) of an event or situation.
• The Preference quantity (P) is a measure of how much the person prefers one circumstance above the other. The defining formula for P is:
  P:= 20 x ¹⁰Log((F+10)/10). The definition implies that negative F-value is emphasized. The reason is that experiencing e.g. a positive F = + 6 (having a nice time) is not something which should be avoided, whereas experiencing a negative value of F = - 6 (having a very lousy time) is something which people want to avoid. (See below).
• Happiness is defined as preference x time. The preference a person experience can change in time. The happiness is the area under the preference graph (an integral).

• (1) T = 40 degrees Celsius and he experiences F = 0 (neutral). P = 2,3.
• (2) T = 44 degrees Celsius and he experiences F = +1 (reasonable warm). P = 0.
• (3) T = 48 degrees Celsius and he experiences F = + 2. (almost ideal temperature). P = 0,7.
• (4) T = 52 degrees Celsius and he experiences F = + 3. (nicely warm). P = 1,1.

This defining formula has another property which is ideally suitable. The fraction of dF to (F + 10); dF/(F + 10) not only decreases is F increases, it grows to minus infinity as F reaches F = -10. From this it follows that F = -10 cannot be attained. Most severe suffering may be around -9.5. Most people experience in their lives probably an F between -8 and + 8. This corresponds to a P = -14 and P = + 5.1 respectively.

Thus:

• (1) Pleasantness-value F in a linear, anti-symmetric scale between -10 and +10.
• (2) Preference-value P calculated from F: P = 20 x (¹⁰Log(F+10) - 1).
• (3) Happiness-value H calculated from P: H = Integral of P in time (continuous case) or H = PxdT of H = P (discrete case)

Moral principles

A moral principle is a relation between the importances which an actor assigns to the resulting happinesses of the receptors. All moral principles accord with the following algorithm:

• Each resulting H-value (H_ij) of action i is multiplied with an importance-weight (w_ji) and the terms are subsequently summed:
  H_i^T:= SUM H_ij x w_ji. A weight is such that w.H < 6 (= 20 x ¹⁰Log((10+10)/10) = 6) and -1 <= w <= +1. This is done with the H-values of each alternative action. The outcome is a number of transformed H-values (H_i^T). If an actor has a dislike for a receptor ánd he doesn't adhere to the ethical principle, then he will weight the H-value of the receptor negatively. If the weight he assigns to an H-value is between 0 and 1, then the actor has an imperfect ability to empathize with his receptor (and he doesn't care or is not interested).
• The maximum of the transformed H-values is chosen: MAX{H_i^T}.

There are 3 standard weight-vectors:

• Ethicalism-vector: w_ji = 1 if and only if ("iff") H_ij has the least value.
• Egoism-vector: w_ji = 1 iff F_ij is the H-value of the actor himself. This is the MAXIMAX egoism or risk-taking (risky) egoism (the maximum of the personal H-maxima of each behaviour alternative). In this text risk-avoiding egoism is not elaborated on. Risk-avoiding egoism is 'ethical'-egoism or ethical principle applied exclusively on the H-values of the actor. This is MAXIMIN-egoism.
• Communalism-vector: w_ji = 1 for all j.

An actor chooses according to a (implicit) personal weight-vector. The vector may depend on the situation. The personal weight-vector can be compared with a (morality) gauge-vector. A (morality) gauge-vector is defined as the three indices that are a measure for how close a weight-vector is to the three standards. Each index is the relative distance the corresponding weight-vector has with respect to one standard-vector. The defining formula for an index is:
1 - [|| (w_1i^Norm, ..., w_ni^Norm) - (w_1i^Actor, ..., w_ni^Actor)||]/||(1,1,1,...,1) - (-1,-1,-1,...,-1)||
The denominator is the length of the maximum difference between two vectors. It can also interpreted as the difference between the maximum difference and the difference between the personal vector and the standard, devided by the maximum difference.

An example. Suppose an actor chooses according to the weight-vector: (-0.5,0.6,0.8). It is given that the three standard vectors are:

• Ethicalism-vector: (0,1,0). ('I want to decrease unheapiness as much as possible.')
• Egoism-vector: (0,0,1) ('Only my own happiness is important. I want to maximize it.')
• Communalism-vector: (1,1,1) ('Everybody is equally important. Somebody who suffers is as important as somebody who has a nice time.')

The ethical principle decreases unheapiness in the world. Suffering is real even though other people can not experience your suffering (if you suffer). Suffering should be avoided as much as possible. (Postulate). Sometimes someone has to suffer more in order to decrease suffering of another one or the person himself in the future. (Analogy: second law of thermodynamics).

The actor chooses the moral principle he applies in the choices he makes. He has a reasoning for this choice. The reasoning is realistic only if it implies that he adheres to the ethical principle. When is an actor motivated to act according to the ethical principle? A sufficient condition is that he thinks objectively. An actor can think objectively or subjectively.

An actor thinks objectively if he realizes/is aware of his own emotional feelings. He has a realistic conception of emotions. The realistic conception of emotions is as follows. Emotion is a special case of feeling. Other feelings are e.g. touch. It is not the same as smell or taste. Emotional feeling is caused (it's physical source) in whole your body, but especially in your stomach. Nerve cells in your stomach send signals to your brains. There they are processed. The processing of the emotion-signals covary with emotion-sensations. The cause of the emotion-sensation that is contained in somebody's mind at each moment in time is not caused directly by somebody else but is mediated by the actor himself. An actor thinks of somebody or sees somebody. This co-varies with emotion-sensations he automatically elicits in himself.

An actor thinks subjectively if he is not aware (metacognition) of the fact that he experiences emotions at the moment he thinks of somebody (or a group of people). If the actor (automatically) elicits in himself an emotion while he is looking at a person or while he thinks of a person and he attributes the virtual property corresponding to the emotion to that person then he thinks subjectively. Example: 'He ís an ´bad person´.' In reality the actor elicits in himself a negative ´bad person´-emotion and attributes the emotion to the person he is looking at as though the person would have an objective ´bad person´-property.

Three examples

An abstract tree diagram example

Three persons are involved in this example; see tree diagram below. E.g. person A can act initially in 2 different ways: action 1.1 (step 1) and action 1.2. (step 1). Behaviour 1.1 has as a result that person A has an H-value +6, person B has an H-value +2 and person C has an H-value 0 (short notation: (+6,+2,0)). Each branch need not be a behaviour. Perhaps in some cases it is useful to let it represent an event that is not a behaviour.

|__ 1.1 __ (+6,+2,0) __ 2.1 __ (0,0,0)
|__ 1.2 __ (+8,+8,0) __ 2.2 __ (-4,-4,-4)
                                |__ 2.3 __ (0,0,0)

Behaviour 1.2 of actor A has as a follow up, 2.2 or 2.3. Each of these two has a probability of occurence, say 0.3 and 0.7 respectively. Each branch from top to bottom has a conditional probability of occurence. The condition is that the actions of actor A (the subject of the diagram) are given (probability = 1). The other actions are from actor B. These actions of B have a certain probability that add up to 1 at each step. E.g. after step 1 in which actor A chooses behaviour 1.2 (that is given), p(2.2) + p(2.3) = 1. Each action 2.2 and 2.3 are of actor B (in step 2).

Each branch from top to bottom has an probable H-value (not to be confuse with H_e, see above). In matrix notation:

	B 2.1	B 2.2	B 2.3
A 1.1	+6,+2, 0	0, 0, 0 (N/A)	0, 0, 0 (N/A)
A 1.2	0, 0, 0 (N/A)	+4, +4, -4	+8, +8, 0

Thus H_prble of branch 1.2 -> 2.2 is: (+8,+8,0) + (-4,-4,-4) = (+4,+4,-4). The conditional probability of occurence is: p(1.2) x p(2.2) = 1 x 0.3 = 0.3.
Thus H_prble = 0.3 x (+4,+4,-4) = (+1.2,+1.2,-1.2) for branch 1.2 -> 2.2.

The H_prble are:

	branch 1.1 -> 2.1	branch 1.2 -> 2.2	branch 1.2 -> 2.3
A 1.1	+6, +2, 0	0, 0, 0 (N/A)	0, 0, 0 (N/A)
A 1.2	0, 0, 0 (N/A)	+1.2, +1.2, -1.2	5.6, 5.6, 0

MIN(A1.1) = 0. MIN(A1.2) = -1.2. MAX{0,-1.2} = 0. Thus in case actor A chooses ethically, he chooses action A 1.1 first. The next choice of A is determined the same way. It depends on what actor B does in step 2.

A mother saves her baby's life.

In this example a baby is in a hazardous situation. The mother of the baby can save her child's life. She has to make stressful effort (H = -4). If she does nothing she will have H_non-emotial = 0 but she disapproves of doing nothing H_non-emotial = - 6 (panic). Thus her H = 0 + -6 = -6. In this model the H-values are: Save: (-4,0) or do nothing: (-6,-8). If the mother chooses according to the egoistic principle, she will choose to save her baby's life. If she chooses according to the communalistic principle, she will choose to save her baby (maximum of {-4+0,-6+-8} = -4 corresponding to saving her baby's life. If she chooses according to the ethical principle, she will also save her baby's live (MAX{-4,-14}= -4). In all three cases this mother will save her baby's life irrespective of whether she is in this situation egoistic, ethical or communal. Her moral attitude causes her to make a stressful effort (in order to save her dearest) because she cares. If she would be cold hearted she would choose doing nothing. Sacrificing yourself such that another person benefits (increasing happiness) from it is altruism. Thus in all three cases the mother is altruistic.

Nuclear Chicken

0, 0	-2, 1
1, -2	-10, -10

After some experience with the same person the actor can learn what the rationality of the other person is and thereby predict his behaviour. There are two possibilities with respect whether an actor has knowledge of the other actor:

• The actor hasn't got the data. In that case he can best play save. If no knowledge then choose A; certain that not outcome -10.
• The actor knows the probabilities of rationalities of groups (probable rationality) given the situation they are in. In that case apply that knowledge to calculate the probable H-values. He thus forms a model of rationality of other actor. But even that is tricky.

• (1) Probability that actor B ethical: 0.1; If A chooses A.1 then actor B will choose: B.1. H₁-vector = (0,0). If A chooses A.2 then actor B will choose: B.1. H₁-vector = (1,-2).
• (2) Probability that actor B egoistic: 0.8; If A chooses A.1 then actor B will choose: B.2. H₂-vector = (-2,1). If A chooses A.2 then actor B will choose: B.1. H₂-vector = (1,-2).
• (3) Probability that actor B communal: 0.1; If A chooses A.1 then actor B will choose: B.1. H₃-vector = (0,0). If A chooses A.2 then actor B will choose: B.1. H₃-vector = (1,-2).