Ethics

by Posted on

Ethics is the philosophy of the judgments of the conduct, character, or circumstances of agential beings living in a society, which judges them to be right or wrong, to be good or bad, or in some similar way, and is used to guide the actions of the agents in the society.

Discussion

Let’s start with the definition of Ethics itself. I’ve been using Lillie’s definition in An Introduction to Ethics (1948) for a semblance of academic credibility:

The normative science of the conduct of human beings living in society, which judges this conduct to be right or wrong, to be good or bad, or in some similar way.

Nitpicks:

  • Does ethics need to be normative?
  • Is ethics necessarily a science?
  • Do the members of the ethics-wielding society need to be human beings and not alien or robot entities?
  • Can ethics judge people or circumstances instead of conduct?

Pros:

  • Doesn’t assume a value system.
  • Includes the context of ethical judgments, namely the society.
  • Notes that ethical judgments can take various forms.

How do we answer these nitpicks?

  • If there is an example of an acknowledged non-normative approach to ethics, then it is indeed not necessarily normative. Metaethical Moral Relativism holds that the weight of moral judgments depends on the traditions and practices of specific groups of persons, for example, being asked to do as the Romans do when in Rome even if it’s perfectly ok to behave otherwise elsewhere.
  • Divine Command Theory stipulates that morality is based upon the commands of God, which seems unscientific, thus either ethics is not necessarily a science or the divine command theory is not actually an ethical theory. Claiming that they’re not doing ethics (right) seems dogmatic and unscientific, too, so this property is probably itself a normative desideratum and not part of the definition proper.
  • See science fiction for compelling arguments that ethics is not limited to the domain of humans; moreover, in this post, I argue that human-centrism in AI Ethics discussions is unethical.
  • One could argue that the primary judgments of consequentialism apply to situations and likewise for virtue ethics’ judgments and character, which have implications for the judgment of actions.

Abstracting these elements, one gets:

The W of the X of the Y beings living in society, which judges this X to be Z.

Where X, Y, and Z are clear.

  • X is the conduct/character/circumstances that are judged.
  • Y is the agential beings that are agents and patients of the judging, also those choosing what to do about said judgments
  • Z is the sort of judgment being made.

W is confusing. Is it appropriate to call ethics a ‘science’? What about a ‘study’? A ‘philosophy’? A ‘body of practices and customs’? The ‘branch of decision theory’, perhaps? The abstraction W almost represents ‘ethics’ itself and should perhaps be dropped as the definitional sentence is reworded.

Ethics refers to the judgments of agential beings living in a society of their conduct, character, or circumstances to be right or wrong, to be good or bad, or in some similar way.

This reworded definition seems to capture the same essence as Lillie’s. The fact that these judgments are then used to guide the behavior of the beings in society is elided; however, one lesson I’ve learned is that statements such as this should probably be additional lemmas and not part of the definition.

The nature of W still poses a challenge because we’d like to say what Ethics is and not simply that Ethics refers to some judgments. Perhaps we can expand it to be the below; however, with formal definitions, it’s probably better to say less than to say too much. Moreover, it’s non-trivial for me to distinguish between “Ethics refers to” and “Ethics is the philosophy of” in SUMO, even if it seems significantly different in English. Some of these simple nuances are surprisingly hard to pin down in a logical ontology.

Ethics is the philosophy of the judgments of the conduct, character, or circumstances of agential beings living in a society, which judges them to be right or wrong, to be good or bad, or in some similar way, and is used to guide the actions of the agents in the society.

SUMO

(documentation Ethics EnglishLanguage "Ethics is the philosophy of the judgments of the conduct, character, or circumstances of 
agential beings living in a society, which judges them to be right or wrong, to be good or bad, or in some similar way, and is 
used to guide the actions of the agents in the society.")
(subclass Ethics Philosophy)

(theoryPhilosophyPairSubclass Ethics MoralTheory)

(=> 
  (and 
    (instance ?MP Ethics)
    (instance ?MT MoralTheory)
    (theoryPhilosophyPair ?MP ?MT))
  (exists (?GROUP)
    (and
      (instance ?GROUP Group)
      (forall (?MEMB)
        (=> 
          (member ?MEMB ?GROUP)
          (instance ?MEMB AutonomousAgent)))
      (forall (?SENT)
        (=> 
          (element ?SENT ?MT)
          (exists (?JUDGE ?BEHAVIOR ?JUDGER)
            (and 
              (instance ?JUDGE MoralJudging)
              (result ?JUDGE ?SENT)
              (subCollection ?JUDGER ?GROUP)
              (agent ?JUDGE ?JUDGER)
              (subclass ?BEHAVIOR AutonomousAgentProcess)
              (refers ?SENT (ClassToSetFn ?BEHAVIOR)))))))))

The current SUMO draft says that every ethical philosophy has a paired ethical theory, which is a syntax-semantics claim. For every instance of an ethical philosophy, there exists a group of autonomous agents where, for each sentence in the moral theory, there is a judgment by some sub-group (possibly a single agent, possibly the whole group as a collective body) of the sentence. This sentence refers to a behavior. The fact that the sentence is a moral sentence implies that it is making a moral value claim.

How to set up the high-level frame is tricky, and the content of SUMO is focused more on low-level details. Probably, the basis of the definition should be the context of an ethics-bearing group. What does it mean for a group of autonomous agents to hold an ethical philosophy? Do we care about ethical philosophies outside of this scope? Theoretically, perhaps.

A perk of the formalization is that it becomes strange to talk about ‘judgments’ sans ‘judger’, which can be tempting when using English. Likewise, the idea that a philosophy is held by a particular entity is also implicit in common human knowledge, whereas when formalizing the definition, we then need to explicitly denote this relationship. Thus, to formalize a ‘simple’ definition, various background theories can be needed that one may not think of immediately. Furthermore, the practical utility of codifying what it means to be a philosophy of some subject is questionable. To do practical ethical reasoning on any particular task, it’s probably not necessary.