1. ETHICS is often said to be the philosophy of morals or ‘theory’ of morals. Questions of morals are to a large extent questions of good and evil and duty. Not all good however is morally relevant and not every duty is a moral duty. With the conception of ethics as the philosophy of morals is sometimes associated a view according to which there is a peculiar moral sense of ‘good’ and ‘duty’ which is the proper object of ethical study.
I shall refer to this view as the idea of the conceptual autonomy of morals. I am referring not so much to a well-defined position as to a certain climate of opinion in moral theory. Nobody I believe has contributed more to the creation of this climate of opinion than Kant. One could therefore also refer to it as a Kantian tradition in ethics.
I have no objection to a definition of ethics as the philosophy of morals. There is not much talk in this book of moral concepts or judgments or rules or principles—and the little which there is is not very systematic The major part of this book does not treat of morals at all. Therefore the tide ‘Ethics’ would not have been well suited for it.
I do however object strongly against the view which I called that of the conceptual autonomy of morals. As I shall try to argue presently moral goodness is not a form of the good on a level with certain other basic forms of it which we are going to distinguish. The so-called moral sense of ‘good’ is a derivative or secondary sense which must be explained in the terms of non-moral uses of the word. Something similar holds true of the moral sense of ‘ought’ and ‘duty’. For this reason it seems to me that a philosophic understanding of morality must be based on a much more comprehensive study of the good (and of the ought) than has been customary in ethics. The name ‘Prolegomena to Ethics’ would not be ill-suited for such a study.
I had thought of using this title. But besides the fact that it has been used before it would be too ambitious. For it is after all only an aspect of the broader approach needed for ethics which I have ventured to study with any thoroughness in the present work. What it is and what the other main aspects are I shall indicate later in this chapter (sect. 4).
2. It has long been current among philosophers to distinguish between normative ethics and ethics which is not normative. Ethics of the first type is supposed to tell what is good and bad and what is our moral duty. Ethics of the second type does not value or prescribe.
The idea of a sharp distinction between ethics which is normative and ethics which is not normative can I think be regarded as an off-shoot of a more general idea of a sharp distinction between norm and fact between the ‘ought’ and the ‘is’. This second idea has become associated in particular with the name of Hume. One could though with caution talk of a Humean tradition in moral philosophy.
The distinctions between the ought and the is and between the two types of ethics is commonly understood in such a way that the term ‘ought’ covers both norms and values and that ‘normative’ as an attribute of ‘ethics’ refers both to the prescriptive and to the evaluative. As another off-shoot of the idea of a sharp distinction between the evaluative and prescriptive on the one hand and the factual on the other hand may be regarded the idea that ‘science’ is value-free (Die Wertfreiheit der Wissenschaften).
On the question what a non-normative study of morals is there is much obscurity and many divergent opinions. Some philosophers particularly from the decades round the turn of the century used to conceive of ethics which is not normative as a science des mœurs i.e. as a sociological and/or psychological study of the ‘natural history’ of moral ideas codes and customs.
There is no doubt a way of studying moral phenomena which is ‘detached’ and ‘scientific’ and which can be sharply distinguished from normative ethics. But it is at least doubtful whether an empirical study of morals is the only form of ethics which is not normative. Many philosophers would deny this. They would maintain that there is a philosophical study of moral concepts and judgments which is distinct both from normative ethics and from the empirical study of moral phenomena. For this type of study of morals the term meta-ethics has recently become fashionable.
On the further question of the nature of meta-ethics opinions are not settled. Some would call meta-ethics a conceptual or logical study of morals. And some would wish to add that a conceptual study of morals is essentially a logical study of the language of morals. Meta-ethics—this seems to be agreed—does not aim at telling what things are good and bad and what are our moral duties. It aims at a better understanding of what ‘good’ and ‘bad’ and ‘duty’ mean.
All these characterizations are loaded with problems. They do not suffice by themselves for drawing a sharp boundary either between meta-ethics and normative ethics or between meta-ethics and empirical investigation.
The idea of a sharp separation of normative ethics and meta-ethics seems to me to rest on an oversimplified and superficial view of the first and on an insufficient understanding of the nature of the second. The view of normative ethics as (some sort of) moral legislation perhaps in combination with a criticism of current moral standards is one-sided. So is the view of normative ethics as casuistry. ‘Normative ethics’ is not a suitable name for any one thing. Those who use the name tend to heap under it a number of different philosophic and moralistic activities. One of these activities thus classified as ‘normative’ I would myself call conceptual investigation; and I would not know how to distinguish it sharply from the allegedly non-normative conceptual analysis belonging to meta-ethics.
Anyone who thinks that a sharp distinction can be maintained between meta-ethics and normative ethics is invited to consider the nature of such works as Aristotle's Nicomachean Ethics Kant's Grundlegung zur Metaphysik der Sitten or John Stuart Mill's Utilitarianism. Is their contents meta-ethics or normative ethics? Some I think would answer that the works mentioned contain elements of both types of ethics and perhaps deplore that their authors did not distinguish more sharply between the two. My own inclination would rather be to say that the difficulties in classification here show the artificiality of the distinction.
3. I would call the investigations conducted in the present work conceptual. I would also agree to saying that the subject-matter of conceptual investigations is the meaning of certain words and expressions—and not the things and states of affairs themselves about which we talk when using those words and expressions. Why is it that I nevertheless do not wish to call the inquiry ‘meta-ethical’ and to regard it as sharply distinguishable from the pursuits of ‘normative ethics’?
My hesitation has to do with the question of the nature of a conceptual investigation. This is a question on which I wish I had and could state a clearer view than I have actually been able to form for myself. I shall here in a brief and somewhat dogmatic way try to state my position with regard to some aspects of it.
First of all I think that there are many types of conceptual investigation many ‘methods’ in philosophy. The choice of a method may depend upon the nature of the problem to be treated or it may perhaps depend upon the temperament of the individual philosopher. I have come to think that the types of conceptual investigation which are best suited for ethics are essentially different from the types suited for theory of knowledge or metaphysics. This difference is probably connected with the fact that the aims as I see them of so-called practical and so-called theoretical philosophy are intrinsically different.
An urge to do conceptual investigations—and one of the main urges to do philosophy I think—is bewilderment concerning the meaning of some words. With the words in question we are usually familiar. We know on the whole how and when to use them. But sometimes we are at a loss as to whether a thing should be called by some such word ‘x’. We are at a loss not because we are ignorant as to whether this thing has some feature y which would be a ground for or against calling it ‘x’. We hesitate because we do not know which features of this thing are grounds for or against calling it ‘x’. We are challenged to reflect on the grounds. Instead of grounds for calling things ‘x’ I could also have said criteria or standards for deciding whether a thing is x or not.
How are grounds or criteria or standards for calling things by words related to meaning? This is a complicated problem on which I shall here only say this much: The meaning of a word has many aspects—and the grounds for calling something by a word I shall call an aspect of the meaning of this word. (If someone wants to distinguish here between criteria and meaning I need not quarrel with him about the meaning of ‘meaning’.)
Reflexion on the grounds for calling things by words is a type of conceptual investigation. How is such investigation conducted? Here a warning is in place. The aim of the type of investigation of which I am speaking is not to ‘uncover’ the existing meaning (or aspect of meaning) of some word or expression veiled as it were behind the bewildering complexities of common usage. The idea of the philosopher as a searcher of meanings should not be coupled with an idea or postulate that the searched entities actually are there—awaiting the vision of the philosopher. If this picture of the philosopher's pursuit were accurate then a conceptual investigation would for all I can see be an empirical inquiry into the actual use of language or the meaning of expressions.
Philosophic reflexion on the grounds for calling a thing ‘x’ is challenged in situations when the grounds have not been fixed when there is no settled opinion as to what the grounds are. The concept still remains to be moulded and therewith its logical connexions with other concepts to be established. The words and expressions the use of which bewilder the philosopher are so to speak in search of a meaning.
I would not wish to maintain that the only fruitful way of dealing with the problems here is to mould the unmoulded meanings to make fixed and sharp that which ordinary usage leaves loose and undetermined. It has seemed to me however that conceptual inquiries which take the form of a moulding or shaping of concepts are particularly suited for the treatment of problems in ethics and some related branches of philosophy (aesthetics political philosophy).
Am I saying that such inquiries aim at stipulative definitions and other proposals concerning the use of language? And is it in the ‘stipulative’ nature of their results that the affinity of these inquiries to ‘normative’ ethics consists?
I do not know exactly how to answer the questions. To say that conceptual investigations sometimes end in stipulative definitions may be true—in some peculiar sense of ‘stipulative’ and in a broad and loose sense of ‘definition’. But to say thus would be on the whole more misleading than illuminating. Here it is good to remember that the philosopher seldom deals with a single concept only. He moves in a field of concepts. This makes him on the whole more interested in logical distinctions and connexions between parts of the field than in the ‘definitions’ of local spots in it.
In ethics conceptual investigations of the type which I have been sketching are a quest for grounds or standards whereby to judge of good and bad and duty. To have such standards is important to our orientation in the world as moral agents. As we shape our standards for judging of good and bad and duty differently we shape the conceptual frame of our moral judgments differently. It does not necessarily follow that the judgments too will be different although they may be. But the grounds on which the judgments are based will be different and therewith their meaning. Our moral ‘points of view’ will be different.
With the remarks in this and the preceding section I have not wanted to deny that there is an activity deserving the name ‘meta-ethics’ and another deserving the name ‘normative ethics’ such that the two are different in kind and sharply distinguishable. But I have wanted to say that there is also a philosophic pursuit deserving the name ‘ethics’ which shares with a common conception of ‘meta-ethics’ the feature of being a conceptual investigation and with a common conception of ‘normative ethics’ the feature of aiming at directing our lives.
4. The concepts which are relevant to ethics may for the purpose of a first approximation be divided into three main groups.
To the first group belong value-concepts. The most important member of this group which is of interest to ethics is the concept good (and its opposites bad and evil).
Concepts of the second group I shall call normative. Here belong in the first place the notions of an obligation a permission a prohibition and a right.
To the third group of concepts belong the notion of a human act and the notions which are relevant to action such as choice deliberation intention motive reason and will. Closely related to them are the notions of desire end need and want.
Concepts of the third group are sometimes called ‘psychological’. They differ however from the psychological concepts studied in epistemology in that they have a ‘total’ character referring to man ‘as a whole’—and not to special faculties such as perception or memory or thinking. Perhaps this is a reason for calling those concepts ‘anthropological’ rather than ‘psychological’ and for labelling their study a Philosophical Theory of Man or Philosophical Anthropology.
Next we have to notice that there are concepts or groups of concepts which fall ‘between’ the three groups and have a ‘foothold’ in two or more of them.
Firstly there are concepts which exhibit affinities both to concepts of the first and to concepts of the second group. Examples are the notions of right and wrong and the idea of justice. The three concepts mentioned can be understood in a ‘legal’ sense which seems to be purely normative. But they can also be understood in a ‘moral’ sense which relates them to ideas of good and evil and therewith makes them value-tinged.
Secondly there are concepts which fall somehow between the first and the third group. For example the notions of pleasure and happiness and their contraries. Is pleasure a value-concept or is it a psychological concept? The question is related to the problem whether pleasantness is a ‘natural’ or a ‘non-natural’ characteristic of things and states.
Thirdly there are concepts which appear to have a foothold in each of the three primary groups. I am thinking chiefly of the generic notion of a virtue and of the various traits of character called virtues—such as courage generosity industry temperance etc. Because of their connexion with the important notion of character they are anthropological (psychological) concepts. They get a normative tinge from their connexion with ideas of a (choice of) right course of action. And finally they are value-tinged due to the connexion between the virtuous and the good and the vicious and the bad man and life.
By a ‘narrow approach’ to ethics I shall understand an approach in the Kantian tradition of the conceptual autonomy of morals. (Cf. sect. 1.) With it I would contrast a ‘broad approach’ to ethics. The latter looks for a place for the moral ideas in the more comprehensive network of concepts of the three basic groups mentioned and the intermediate groups.
One could perhaps distinguish between two main directions which a broad approach to ethics might take. One can look for a foundation of morals so to speak in a ‘vertical’ dimension in the needs and wants of man and in the specific nature of man as agent. On this view a clarification of the concepts of the third group above the ‘anthropological’ or ‘psychological’ ones is the most urgent preliminary to moral philosophy. Ethics one could also say has to be set in the perspective of a Philosophical Anthropology.
One can however also look for a foundation of morals in a ‘horizontal’ dimension trying to place the moral notions in a broader setting of value-concepts and normative ideas. Ethics then as it were emerges as a special branch of a General Theory of Value and Norm.
I do not wish to maintain that the two methods are exclusive of one another or that they can be sharply distinguished. Nor would I say that one of them must be given priority if one cannot pursue both. The first may be regarded as the more fundamental in the sense that one cannot study norms and values in isolation from the psychological concepts whereas one can study the latter in relative isolation from the former.
If the method adopted in this treatise has to be placed in the above typology we shall have to say that it is of the ‘horizontal’ rather than of the ‘vertical’ type. It is furthermore an approach to ethics from the side of value rather than from the side of norm. (This last has to do with the view which I tend to take on the mutual relations between norms and values. Cf. Chapter VIII.) More specifically still: it is an approach to ethics from a study of the concept of goodness.
It may be true as has been said that a better understanding of the concept of a human act and of the related psychological concepts is needed before we can successfully tackle the problems of ethics. But it seems to me equally true that a necessary preliminary to a successful study of moral action moral goodness and moral duty is a study of goodness in all its varieties. I do not know that this study had been systematically undertaken before. A beginning to it is attempted in the present work.
5. By the Varieties of Goodness I understand the multiplicity of uses of the word ‘good’. A useful preliminary to the study of this multiplicity is to compile a list of familiar uses of the word and try to group them under some main headings.
We speak of a good knife watch hammer razor and other artefacts which are used as instruments for various purposes. We also speak of good dogs cows horses and other animals whom man has domesticated for his needs and ends. In this same region belong a good car a good house a good harbour. Hither also belong all that can be called ‘a good way of doing’ something e.g. unlocking a door making one's bed or memorizing a poem.
I shall group all these uses of ‘good’ under the head instrumental goodness. It is not maintained that whenever ‘good’ is used as an attribute of say ‘dog’ or ‘house’ it connotes instrumental goodness. Which form of goodness is meant cannot be seen from the phrase alone but must be gathered from its use in a context.
We further talk of a good chess-player runner orator driver general business-manager administrator carpenter scientist and artist. A common characteristic of such men is that they are good at something. That at which they are good is usually some activity or art for which a man may possess a natural talent but in which he will also have to undergo some special training before he can excel in it. I shall coin for this excellence the name technical goodness.
There are borderline cases between instrumental and technical goodness for example the use of ‘good’ in ‘a good salesman’ or ‘a good servant’ or ‘a good soldier’. The goodness of a good achievement or performance is often of the form which is here called technical.
As constituting a group of their own I shall regard the uses of ‘good’ as an attribute of organs of the body and faculties of the mind: for example when we speak of a (in the medical sense) good heart of good eyes good sight good memory. I can suggest no better name for this form of the good than medical goodness. It is related to and yet characteristically different from technical goodness. One of the differences comes to light in the relationship which the goodness and badness of organs and faculties have to the notions of health and illness—and through them to the more comprehensive notion which I propose to call the good of a being or welfare.
Medicine is good for the sick exercise for the health manure for the soil lubrication for the engine to have good institutions is good for a country good habits for everybody. Generally speaking something is said to be good for a being when the doing or having or happening of this thing affects the good of that being favourably. For this form of goodness I shall here reserve the name of the beneficial. That again which affects the good of a being adversely is called harmful damaging injurious ‘a bad thing’ or quite often an evil.
The beneficial is a sub-category of the useful or of that which I shall also call utilitarian goodness. Related to the useful are the advantageous and the favourable. When we speak of a good plan a good opportunity good advice good luck good news we are usually thinking of something which is useful or advantageous for some purpose or pursuit. The beneficial could be distinguished from the ‘merely useful’ by saying that the useful is that which favours some end or purpose in general and beneficial that which promotes the special end which is the good of some being. The good of a being however is not only a very special end. It is an ‘end’ only in a special sense of the word. The idea of the good of a being as an end as far as I can see is the same as the idea entertained by many philosophers of (some) beings as ends in themselves.
In English the substantive ‘good’ has at least three meanings which it is important to keep apart. By ‘the good’ we can understand goodness the concept or idea of good. By ‘the good’ we can also understand that which we have called the good of a being and for which ‘welfare’ is another name. Finally there is the substantive ‘good’ which has the plural ‘goods’. In German these three meanings of the substantive are very handily distinguished by means of the three words das Gute das Wohl and das Gut (pl. die Güter).
By ‘a good’ one can understand anything which is a bearer of the value ‘good’ in short: anything which is good. This is typically a philosopher's use of the term. I shall not adopt it here. By ‘a good’ however one can also understand anything which is an end of action or object of desire or want or need. When in this treatise I sometimes use ‘good-goods’ as a pair of technical terms I use it in the second of the two senses mentioned. Extensionally the two senses as easily noted are largely overlapping.
An important form of goodness which has so far not been mentioned at all is the hedonic good. We speak of a good smell or taste a good apple or wine a good dinner a good joke a good holiday or time good company good weather. Not always when ‘good’ is used in such phrases as those just enumerated does it stand for that which is here called hedonic goodness. Good weather for example is not necessarily agreeable weather which is a hedonic feature; it can also be favourable say for the harvest which is a utilitarian quality.
As the useful and its sub-form the beneficial is related to that which a man wants and needs in a similar manner the hedonic good is related to that which a man enjoys and likes—to get to have to do to be. The hedonic moreover has many rather different sub-forms. Sometimes the hedonic is near the aesthetic sphere. What a good smell is to the nose a nice tune is to the ear and a beautiful sight to the eye. The hedonically good is frequently called amusing or entertaining sometimes refreshing or exhilarating. It can always I think be called pleasant. The philosopher's traditional term for it is as well known pleasure.
Further there are the uses of ‘good’ which refer to matters of conduct and character. A man can be good and do good. Good is usually done to somebody. When a man does good to some being his acts or deeds are frequently though not always called good as well. An act can be done from a good motive or with a good intention. There is a feature of character called benevolence and a preparedness to act called a good will. Of this last one philosopher said that it is the only thing in the world which is ‘good without qualification and restriction’. It is the good in matters of conduct and character which is above all related to the so-called moral life of man.
The enumeration and grouping of uses of ‘good’ which we have given is very far from exhaustive. Examples could easily be given of uses which fall either completely outside any of the forms which we have here tentatively distinguished or which seem to fall somewhere between them. The reader is invited to consider whether the following typical phrases with the word ‘good’ in them should be classified with our above forms or where he would wish to place them on the logical map: ‘good manners’ ‘good times’ ‘good incomes’ ‘good sleep’ ‘good appetite’ ‘good-tempered’ ‘a good conscience’ ‘a good reputation’ ‘good as gold’ ‘a good specimen (sample example)’ ‘a good impression’ ‘a good view’ ‘a good reason’ ‘a good idea’ ‘a good book (work of art)’ ‘a good wife’.
What has been said in this section should give the reader an impression of the semantic multiplicity and logical wealth of the phenomenon which I have called the Varieties of Goodness It should also show the inadequacy and artificiality of such schematisms as say the traditional classification of all good—sometimes of all value whatsoever—into two main types viz. good as a means and good as an end instrumental and terminal extrinsic and intrinsic good. We shall make no use of these dichotomies for our present purposes. But I think it is appropriate to warn us against them once and for all.
6. Is there a unity among the Varieties of Goodness? If there is a unity what is its nature? If there is nothing which unites the various forms of the good how is it that the word ‘good’ has come to have this multiplicity of uses?
One way to achieve unity in the variety here would be to view the forms of goodness as species of a generic good. This possibility would imply two things. Firstly that there is some feature which is common to all forms of goodness something in which they all share or participate. Secondly that there are distinguishing features which when added to or combined with the generic feature mark off the various forms of goodness from one another. The generic feature would explain why instrumentally technically etc. good things are good. The specific differences would explain why some good things are instrumentally some technically etc. good.
Man and dog are both species of the genus mammal. Individual men and individual dogs all have the essential characteristics of mammals—‘the form of mammalhood’—in common. These features ‘make’ them mammals. Other specific features distinguish them as men and as dogs. Is the relation of the concepts of man and dog to the concept of mammal the logical pattern for understanding the relation of the forms of goodness to the idea of good?
It is against an affirmative answer to this question that Aristotle if I understand him rightly is polemical in the sixth chapter of the first book of the Nicomachean Ethics. In opposition to Plato Aristotle did not acknowledge the existence of an Idea of Good. One of his reasons for not acknowledging this seems to have been his opinion that the things called ‘good’ did not compose a genus by virtue of a common characteristic their goodness. But the exact nature of Aristotle's own view of that which could conveniently be called the meaning-pattern of the word ‘good’ is not at all clear to me.
One could distinguish between the Varieties of Goods when ‘goods’ means ‘good things’ (see above p. 10) and the varieties of Goodness. That both wines and carpenters and lungs can be good is an example of how diverse good things can be. That ‘good’ sometimes means ‘pleasant to taste’ sometimes ‘skilful’ and sometimes ‘healthy’ is an example of how varied are the forms of goodness. Aristotle did not distinguish clearly between the two sorts of variety. His argument against the Idea of Good draws support from the Varieties of Goods rather than from the Varieties of Goodness. It is obvious however that the former variety is secondary to the latter. It is not because the good things are such a mixed bunch that there are so many forms of goodness; but it is because of the multiform nature of goodness that things of the greatest dissimilarity in kind and category can be good. An argument against the view that there exists a generic good must therefore be based on considerations pertaining to the Varieties of Goodness rather than the Varieties of Goods.
It seems to me certain that the forms of goodness are not related to a generic good as species are related to a genus. (This is why I speak of ‘forms’ and not of ‘kinds’ or ‘species’ of goodness.) But I do not know how to argue conclusively for my opinion.
No attempt will be made in this work to make clear the notion of a form which I use when speaking of the forms of goodness. The relation of a form of X to X is not that of species to genus nor that of occurrence to disposition nor that of token to type nor that of individual to universal. Which the relation is we shall not discuss.
Of the forms of goodness I sometimes distinguish sub-forms. Thus for example three sub-forms of technical and three of hedonic goodness will be mentioned. It is not maintained that the relation between a sub-form of a form of goodness and this form of goodness were the same as the relation between a form of goodness and goodness itself. The nature of the relation of sub-forms to forms is another question which will not be discussed.
7. It is sometimes said that the word ‘good’ is ‘vague and ambiguous’. Could the notions of vagueness and of ambiguity be used for throwing light upon the phenomenon which we have called the Varieties of Goodness?
It is important to distinguish between vagueness and ambiguity. The first is a feature of some concepts or as one could also say here meanings. The second is a characteristic of some words (and phrases). For this reason to refer to something as being ‘vague and ambiguous’ is a rather confused mode of speech. But the two suggestions that the meaning of ‘good’ is vague and that the word ‘good’ is ambiguous are worth serious consideration.
I shall say that a concept is vague when there easily occurs uncertainty whether a thing should be classified as falling under the concept or not and when this uncertainty cannot be removed by appealing either to (further) facts about the case or to existing criteria for the application of the concept. Vagueness in this sense is highly characteristic of some of the forms of the good which we have distinguished chiefly I think of instrumental and technical goodness. For example: whether a carving-knife is good or not for its proper purpose may be difficult to decide because of vagueness and similarly whether a man is a good wrestler or not. But it is not because of a vagueness attaching to the idea of good that both a knife and a wrestler are capable of falling under the concept of goodness. In other words the fact that some of the typical uses of ‘good’ are vague cannot account for the fact that there are these many uses (forms of goodness). The notion of vagueness therefore is not relevant to an understanding of the phenomenon which we here call the Varieties of Goodness.
Ambiguity may be defined as a (logically) accidental identity of words standing for different ideas.1 The suggestion that the Varieties of Goodness were due to an ambiguity of the word would amount to saying some such thing as that ‘good’ in ‘good knife’ and in ‘good intention’ mean two different things which by accident only are called by the same name.
When a word for a certain concept in a language is ambiguous then it is pure coincidence if the word for that same concept in another language is in the same way ambiguous too. If ‘good’ in English really were an ambiguous word we must wonder at the fact that bon in French dobrij in Russian and agathos in Greek also happen to exhibit the same ambiguities. The range of things and kinds of thing called ‘good’ in English are much the same as the things and kinds of thing called dobrij in Russian. This observation alone can be considered decisive evidence against the suggestion that ambiguity was responsible for the existence of the many forms of goodness.
Just as ambiguity must be distinguished from vagueness it must also not be confused with analogy. ‘Port’ in English is ambiguous. The etymological fact that the name of the wine derives from the place-name ‘Oporto’ which in its turn is derived from the same root as the English word for harbour does not imply an affinity between the two meanings of ‘port’. The word ‘deep’ provides a good example of analogy. In ‘a deep well’ the word is used with that which I shall call its primary or literal meaning. In ‘a deep thought’ it is used with a secondary or analogical or metaphorical meaning. Analogies are on the whole intra-linguistic. This indicates that they are grounded in affinities between meanings and not as ambiguities are on linguistic accidents. It is natural to think of the Varieties of Goodness as a variety of meaning-affinities. The idea that the Varieties of Goodness were due to analogy therefore has a certain plausibility.
It seems to me however that there are decisive arguments against the idea that the diversity of the forms of the good were a diversity of analogical meanings of the word ‘good’. Analogy presupposes a primary use of the word with which we must be familiar before we can understand and make use of the analogy. When a word has analogical meanings it is usually in the individual case clear whether the word is being used in its primary sense or with an analogical meaning. It is clear that the Ocean is deep literally and a thought ‘only’ metaphorically that a face is smiling literally and a corn-field metaphorically. Nobody would maintain the reverse here. But is there a literal sense of ‘good’ of which the other senses are analogical or metaphorical extensions? Is e.g. instrumental goodness primary and hedonic secondary or vice versa? Or are not both equally basic (‘literal’)? Is it only by analogy to a good apple and a good knife that we speak of a good act or a good general?
It is among other things the obvious difficulty of singling out one (or some) of the uses of ‘good’ as primary or literal which speaks strongly against the idea that the variety of forms of goodness is a variety of primary and analogical meanings of the word ‘good’. With this I have not wished to exclude the possibility that there actually are some analogical uses of ‘good’. Some such analogical uses may even be philosophically of great interest. Perhaps goodness as an attribute of a supreme being (God) is analogical.
Wittgenstein in his later writings used the idea of what he called family-resemblance. The idea is related both to vagueness ambiguity and analogy and yet different from them all. Family-resemblance gives unity to such concepts as those of language or sentence or number or—to use one of Wittgenstein's favourite examples—the concept of a game. We may call them family-concepts. The philosophic importance—as I see it—of the idea of family-resemblance is that the insight into the family-character of a concept may make us give up an attempt to hunt for its ‘essence’ i.e. for a common feature of all things falling under this concept which would explain to us why these things are classified together.
There is some suggestion in Wittgenstein's writings that goodness were a family-concept that all uses of ‘good’ formed a family of cases. There may be some truth in this. But I doubt whether it is a useful suggestion on the whole. The question ‘What is good?’ even when understood as a conceptual and not as an axiological question seems to me very unlike the question ‘What is language?’ or to take a question bordering on problems of value ‘What is art?’. The notion of art is I think typically a family-concept. Many problems of philosophical aesthetics incidentally seem to me relevantly connected with this character of concepts.
Often symptomatic of the family-nature of a concept is a bewilderment as to whether something ‘really’ falls under this concept. Is photography really an art? Or the drawing of cartoons? Is the language of bees really a language? Such questions can express a genuine philosophic puzzlement. If goodness were a family-concept like art we should expect bewilderment as to whether some of the forms of goodness really are forms of goodness. But such puzzlement is apparently not felt. We may feel astonished at the fact that a smell as well as a general rainfall as well as lungs can be good. It may that is be astonishing to consider how diverse good things can be. But we should not hesitate to call the pleasant or the skilful or the useful or the healthy forms of the good.
Another difference between goodness and some typical family-concepts seems to be this: New members of a family may originate in the course of history. New games are invented new forms of linguistic communication are created or can be imagined also maybe new forms of art. But the forms of goodness do not seem to be in the same way relatable to temporal changes. Conceptual observations may lead a logician or philosopher to distinguish between uses of ‘good’ which had before been classified together and regard them as separate forms or sub-forms of goodness. But the forms or sub-forms thus distinguished would not be new inventions but familiar phenomena among which a new difference was noted.
Observations of the sort which we have here mentioned seem to me to speak against the idea that goodness were typically a family-concept.
What I have ventured to say in this and the preceding section about the Varieties of Goodness are essentially negative things. The unity in the variety if there is one is not that which a genus gives to the species falling under it. Nor does it appear to be a unity of the sort for which analogy or family-resemblance can be held responsible. Ambiguity and vagueness again do not account for the variety of the forms of the good. The meaning-pattern of ‘good’ is peculiar and puzzling. It is worth more attention than it has received on the part of philosophic semanticists.2 It is not however my plan to discuss these aspects of the Varieties of Goodness in the present work. I have only wanted to draw attention to some problems in the region.
8. From the problem of the nature of and reasons for the variety of the forms of goodness one may distinguish the problem of the logical relationships between the forms. The two problems are not unrelated. If it could be shown that there is one basic form of the good in the terms of which all the others can be defined then the solution to the problem of relationships between the forms would automatically cater also for the problem of the semantic character of their variety. The same would be the case if it turned out that all forms of the good can be defined in the terms of a small number of more basic notions.
There will be no systematic discussion at all in this work of the question of the definability of ‘good’; nor of the related problem whether ‘good’ is the name of a ‘natural’ or ‘non-natural’ characteristic. Occasionally some observations will be made which are relevant to these issues. ‘Good’ has a number of partial synonyms which can replace it in some contexts without change of meaning. ‘Pleasant’ ‘skilful’ ‘healthy’ ‘useful’ are such words. That which we are going to say of their relation to ‘good’ will not however bear directly on the question of relationships between the forms nor will it contribute to an explanation of their variety.
Certain affinities between the forms of goodness are striking. One is the relation between that which I have called instrumental and technical goodness. Excellence in a skill is technical goodness. On account of their skills men serve various purposes of society. This gives to their skills an aspect of instrumental goodness. There is further a relationship between technical and medical goodness. Good organs like good artisans and professionals do their proper job well. The proper job of the organs is to serve the good of the body in the last resort the good of man.
Instrumental and utilitarian goodness are closely related although not to the point of being indistinguishable. Sometimes instrumental goodness can be called a degree of usefulness (utilitarian goodness). A sub-form of utilitarian goodness is the beneficial. It is related to the notion of the good of a being. In this respect it resembles medical goodness.
The relationships mentioned between forms of the good do not as far as I can see open up a possibility of defining some of the forms in the terms of some other forms. The notion of moral goodness however holds a peculiar position in this regard. It craves for a definition. The fact that there has been so much dispute about its nature and relatively little about the nature of the other forms of the good may be regarded as symptomatic of this very craving. Some philosophers would perhaps regard the craving as satisfiable only through the insight that moral goodness is not definable but sui generis an irreducible form of the good. I would hold the exact opposite of this view. I shall later (Chapter VI) propose an account of moral goodness (as an attribute of acts and intentions) which defines it in terms of the beneficial. Moral goodness is thus not logically on a level with the other forms which we have distinguished. It is a ‘secondary’ form. I shall prefer not to talk of it as a special form of goodness at all.