Rawls J. Lectures on the History of Political Philosophy

Подождите немного. Документ загружается.

edition (Methods of Ethics, p. vii; hereafter ME) that he aims to examine (and

I would add compare and contrast) all “the different methods of obtaining

reasoned [moral] convictions as to what ought to be done which are to be

found—either explicit or implicit—in the moral consciousness of mankind

generally.” These methods “have been developed, either singly or in combi-

nation, by individual thinkers, and worked up into the systems now histori-

cal” (p. vii). Sidgwick wishes to describe and criticize (assess) these methods

“from a neutral position, and as impartially as possible” (p. viii). Part of our

task here is to see what this neutral and impartial position is.

What is a “Method of Ethics”? Sidgwick deﬁnes it as any rational proce-

dure by which we determine what individual human beings ought to do; or

determine what is right for them to do; or to seek to realize by (free) volun-

tary action (ME, p. 1). The phrase “individual human beings” distinguishes

ethics from politics, which Sidgwick says studies what is right or good legis-

lation,

but this distinction is not important for us, since the principle of

utility applies to both and Sidgwick’s discussion of justice really belongs to

politics.

Note that Sidgwick assumes that, under any given circumstances, there

is something (some deﬁnite alternate institution or custom, etc.) that it is

right or reasonable to do, or to bring about (if this is possible); and that this

may be known in principle. (See ME, Preface, 1st edition, p. vii.) In addition,

Sidgwick assumes that a rational method is one that can be applied to all ra-

tional (and reasonable) human beings to get the same result, when the

method is correctly followed (cf. ME, 27, 33). In sum: There is always one

right or best answer, and this answer is the same for all rational minds. For

Sidgwick this assumption is characteristic of science and the search for

truth; he believes it holds for moral philosophy and ethical beliefs. Sidgwick

says: it is implied in “. . . the very notion of Truth that it is essentially the

same for all minds, [and so] the denial by another of a proposition that I

have afﬁrmed has a tendency to impair my conﬁdence in its validity” (ME,

p. 341).

This is said in explaining why a mark of self-evidence is general agree-

ment in judgment. Thus Sidgwick holds the thesis of moral objectivity.

(2) Now the methods of ethics Sidgwick has in mind are those proce-

[ 381 ]

Four Lectures on Henry Sidgwick

6. “Politics . . . seeks to determine the proper constitution and the right public con-

duct of governed societies.” ME, p. 1.

EXAM COPY

dures embedded in the historical doctrines: the various forms of rational

intuitionism and moral sense views; perfectionism and utilitarianism; and

social contract doctrines insofar as they incorporate parts of such doctrines.

Sidgwick also includes rational egoism as a method of ethics.

Note that Sidgwick wishes to concentrate on the methods themselves

and their differences as methods, and not on their practical results. He

wants to put aside the desire to edify which he thinks is a barrier to advance

in ethics, and to study the methods from disinterested curiosity. He wants

to forget about even “ﬁnding and adopting the true method of determining

what we ought to do; and to consider simply what conclusions will be ra-

tionally reached if we start with certain ethical premises, and with what de-

gree of certainty and precision” (ME, Preface, 1st edition, p. viii).

This statement does not quite describe Sidgwick’s view accurately, since

he is prepared to say that a rational method of ethics must answer to certain

criteria; and these criteria serve, as we shall see, as the neutral standpoint

from which the different methods can be assessed. Nevertheless, the desire

to set out and to compare the various methods of ethics, from an impartial

point of view, is an important feature of Methods of Ethics.

The implication of this feature is that we should not view Methods of

Ethics as aiming to justify classical utilitarianism. It is plainly the doctrine

Sidgwick prefers and the one to which he is most strongly drawn. But at the

end of ME he thinks he is forced to recognize that while utilitarianism,

from a neutral point of view, passes the criteria of a rational method of eth-

ics far better than any form of intuitionism, and so is superior to

intuitionism, nevertheless classical utilitarianism and rational egoism both

seem to pass these standards equally well. Sidgwick comes to the unwel-

come conclusion that there seems to be a conﬂict of reason within itself in

the practical sphere.

(3) The structure of ME is roughly as follows (I go over this so that you

can put Ch. 5 on Justice and Sidgwick’s argument as a whole in its proper

context):

(a) ME falls into 4 Books.

Book I: discusses preliminary matters: deﬁnitions of ethics and of moral

judgment, ethical principles and methods; deﬁnition of free will and its re-

lation to ethics; deﬁnitions of desire and of pleasure; of intuitionism vs.

egoism and self-love, etc.

Book II—Egoism: since Sidgwick decides that there are essentially only

[ 382 ]

appendix

EXAM COPY

three fundamentally distinct methods of ethics, rational egoism, intuitionism,

and utilitarianism, he sets out to give a systematic comparison and descrip-

tion of these. Book II is devoted to rational egoism.

Book III—Intuitionism: covers the various kinds of intuitionism (to-

gether with Chapter 8 of Book I), and at the same time points out the

weakness of intuitionism as a method and hints at the argument to come,

that classical utilitarianism is superior. See especially Book III, Chapter 11,

on the review of common-sense morality, then Chapter 13 on Philosophical

Intuitionism, and Chapter 14 on Ultimate Good.

Book IV—Utilitarianism: begins with the deﬁnition of the principle of

utility in its classical form. Chapter 1 presents part of the neutral and im-

partial point of view, or argument whereby methods of ethics can be as-

sessed. Chapter 2 discusses the proof of the principle of utility; Chapter 3

surveys the relation between common sense and utilitarianism and argues

that common sense is, as it were, unconsciously utilitarian. Chapters 4–5

lay out the method of utilitarianism; and Chapter 6 discusses the relations

between the three methods of ethics and closes with the dilemma of the

“dualism of practical reason.”

(b) Strictly speaking, the argument of Methods of Ethics does not justify

the classical utilitarian doctrine, although it is clearly the view to which

Sidgwick is strongly inclined. The reason is that while utilitarianism wins

out over intuitionism in Bks. III–IV, a tie exists between utilitarianism and

rational egoism: that is, both satisfy equally well the objective criteria of a

rational method of ethics. This offhand surprising conclusion is reached in

the last chapter of Bk. IV; so we have, Sidgwick says, the dualism of practi-

cal reason and no objective resolution is in sight.

Thus, from this structure and outline it is clear that Sidgwick fails in his

objective: While he is satisﬁed, for the moment, that he has described and

compared the main methods of ethics correctly, it turns out that at least

two of them—rational egoism and utilitarianism—pass equally, so far as he

can tell, the rational and neutral tests of any such method. Therefore his

initial assumption of objectivity—the thesis that there is always one right an-

swer—is put in question. He suggests a way out via a theological assump-

tion, but this we have not time to examine (I believe it well worth looking

at, however convinced one may be that it cannot be correct).

(4) I should mention here (it will be relevant later), ﬁrst, that Sidgwick

reduces the main methods of ethics to three only—not, to be sure, without

[ 383 ]

Four Lectures on Henry Sidgwick

EXAM COPY

examining the others of historical importance: rational egoism (Bk. II);

intuitionism (Bk. III); and utilitarianism (Bk. IV). Thus perfectionism re-

duces to intuitionism; and Kant’s doctrine reduces to a formal principle of

equity or fairness in Sidgwick’s terminology (cf. ME, p. 379). This is, in my

opinion, too small a range of comparisons: it is defective, I think, in not see-

ing that Kant’s doctrine, or a view similar to it, is a distinctive method of

ethics; and Theory of Justice is such a view. Also, I think he wrongly assimi-

lates perfectionism to intuitionism. This gap in Sidgwick’s range of compar-

isons is one weakness of his overall view.

Secondly, I believe Sidgwick fails to include in his description of meth-

ods of ethics certain important aspects of a moral conception, but I shan’t

go into this at this point.

(5) The General Criteria of any Rational Method of Ethics:

I call your attention to the footnote, p. 293 (of Ch. 5) where Sidgwick

says that by “arbitrary” (as applied to deﬁnitions) he means deﬁnitions that

include limitations (exceptions and qualiﬁcations) “as destroy the self-evi-

dence of the principle; and, when closely examined, lead us to regard it as

subordinate.” Now in the background here is Sidgwick’s view of the criteria

for the 1st principles of a Rational Method of Ethics, which are:

(a) First, Sidgwick holds that the 1st principles of a method of ethics

must satisfy these conditions: (i) they must be at least as certain as any other

moral principles, and (ii) of superior validity to other principles; (iii) they

must be really self-evident and derive their validity, or evidence, from no

other principles, moreover:

(b) Any such principles (iv) must be fully rational in the sense that they

contain no limitations, or exceptions, or restrictions, unless these are self-

imposed; that is, follow from the principle itself, and are not simply ap-

pended as unexplained provisos (cf. Methods of Ethics: 293n., the deﬁnition

of “arbitrary”). In addition:

principles and standards (and lower-level moral precepts and beliefs) so as

to organize them into a complete and harmonious scheme free of arbitrary el-

ements. This requirement is connected to another: namely that (vi) 1st

principles must deﬁne a method of ethics that determines (ascertains) ac-

[ 384 ]

appendix

7. Here I interpret Sidgwick’s procedure and arguments, but see especially Book IV: 2.

8. Here you should see Methods, III, Ch. 11 and IV, Ch. 2; and Schneewind, Chs. 9–10.

EXAM COPY

tual rightness, and not merely prima facie rightness—1st principles must

yield a correct judgment, all things considered; and so (vii) must serve for

rational agents as an actual guide to practice, and so enable us to act ratio-

nally—hence ﬁrst principles cannot be vague, imprecise, and ambiguous;

and ﬁnally (viii) a ﬁrst principle must be one that suitably corrects our pre-

reﬂective judgments.

Sidgwick’s account of justice (Bk. III, Ch. 5) is designed to show that

none of the principles of justice found in common sense meet these criteria

and hence are subordinate principles. It is particularly the last three condi-

tions, (d)–(f), that he argues throughout Bk. III, Ch. 5, on Justice, although

the ﬁrst three are there also. We turn to Sidgwick’s account of justice in the

next lecture.

lecture ii

Sidgwick on Justice and on the Classical Principle of Utility

§1. Sidgwick’s Account of Justice

(1) You should read Sidgwick’s account of justice in Book III, Chapter 5,

as part of his long and careful account of the intuitive principles found in

common sense, and reﬁned by various writers in the effort to formulate

them as bona ﬁde and rational ﬁrst principles. He believes that his survey

of these principles shows that in every case, these principles prove to be

vague and imprecise once we try to apply them in practice; and that they

are subject to various exceptions and qualiﬁcations that are arbitrary in the

sense that the principles themselves do not include any explanation of the

rational basis of these exceptions and qualiﬁcations. Therefore, Sidgwick

concludes, these principles cannot be bona ﬁde rational and objective ﬁrst

principles. There must be some other and higher controlling principle or

principles that accounts for these qualiﬁcations and provisos. And he hints

throughout (and often more than hints) that this higher principle must be

the principle of utility. All this, of course, on the assumption that there is al-

ways a right or true answer, and that we can know it and agree upon it (if

we follow reason).

[ 385 ]

Four Lectures on Henry Sidgwick

EXAM COPY

(2) Sidgwick discusses the notion of justice on three occasions: the full-

est is that in Methods, Ch. 5 of Book III; the next is the brief summary of

Ch. 5 that Sidgwick gives in his “Review of Common Sense,” Ch. 11 of

Book III, at 349–352; and ﬁnally there is the assessment in Ch. 3 of Bk. IV, at

440–448.

Sidgwick explains the different aims of these discussions as follows. In

Ch. 5 of Book III the aim is “to ascertain impartially what the deliverances

of Common Sense actually are” (ME, p. 343); while the aim in the “Review” of

Ch. 11 is “to ask how far these enunciations [i.e. the deliverances of common

sense] can be claimed to be classed as Intuitive Truths” (ME, p. 343). In Ch. 3

of Book III the aim is to show that in coping with the difﬁculties and ambi-

guities, etc., that arise in practice in deﬁning and specifying its notions of

justice, common sense is, as it were, unconsciously utilitarian, since the prin-

ciple of utility is naturally invoked, even if only implicitly. (One of Sidgwick’s

deﬁnitions of common sense of mankind is this: what is “expressed gener-

ally by the body of persons on whose moral judgments [one] is prepared to

rely”; ME, p. 343). So while these various accounts of justice are somewhat

repetitive, their stated aim is different; and in fact Sidgwick’s observations

are not the same and they supplement each other to some extent.

(3) A rough outline of Ch. 5 is as follows:

(a) In §1 (ME, pp. 264–268) Sidgwick holds that while justice is con-

nected in our minds with laws (cf. administration of justice), it cannot be

identiﬁed with what is legal, since laws may be unjust. Again, while justice

includes and implies the absence of arbitrary inequalities in framing and ad-

ministering laws, it is not merely this either.

(b) In §2 (pp. 268–271) Sidgwick discusses what he calls “conservative jus-

tice,” that is: the fulﬁllment of (1) contracts and deﬁnite understandings,

and (2) expectations that arise naturally out of the established practices and

institutions of society. However, the duty of fulﬁlling the latter is not

clearly deﬁned; nor is it clear how much weight these expectations should

have.

as judged by the standard of ideal justice. But there are different conceptions

of this standard.

(d) In (§4) (pp. 274–278): One view is that freedom is the absolute end;

but the attempt to elaborate an ideal notion on this basis runs into insuper-

able difﬁculties.

[ 386 ]

appendix

EXAM COPY

[ 387 ]

Figure 8. Schema of Sidgwick’s account of justice, Methods, Bk. III, Ch. 5

EXAM COPY

(e) In §5 (pp. 278–283): Nor does the realization of freedom answer to

our common conception of ideal justice, which is rather that desert should

be requited.

(f ) In §6 (pp. 283–290): But the application of this principle likewise is

very perplexing for it admits of different interpretations of desert: for ex-

ample, desert may be estimated by conscientious effort, or by worth of

what is done (of services); moreover, the principle of ﬁtness is a complicat-

ing factor.

(g) §7 (pp. 290–294): Similarly, there are difﬁculties with ill desert in de-

ﬁning criminal justice. Sidgwick ends with a summary of his conclusions

(pp. 293–294).

§2. Statement of the Classical Principle of Utility

Intuitively, the idea is to maximize the net balance of pleasure over pain.

(1) The principle applies quite generally to all subjects: in situations and

practices, individual actions and traits of character, etc., and in all circum-

stances, both ideal and non-ideal. Thus: in any given situation, that institu-

tion or action, etc., is right, or that which ought to be done, if among all

the feasible alternatives realizable in the circumstances, it is the one that

maximizes:

au au au au

ii nn

=+++

∑

11 22

...

(a linear sum of the u

’s)

where the a

’s are real numbers (the weights of the u

’s) and the u

’s are real

numbers that represent the utility (the net balance of pleasure over pain)

for each individual I, these numbers taking into account all the conse-

quences of the institution or action in question on every one of the individ-

uals affected whatever their position in space or time, and so, e.g., however

far into the future.

(2) To ﬁx ideas assume that the individuals in question belong to the

same society and leave aside all other individuals; however, include all persons

for m generations into the future, when by hypothesis, the world comes to

an end. The idea is the maximize utility over this stretch of time, leaving

out the past, since bygones are bygones and not affected by human action.

(3) In the classical doctrine the weights a

all = 1, for as J. S. Mill says

this is implied by the notion of measuring pleasures and pains as objective

[ 388 ]

appendix

EXAM COPY

quantities given by their intensity and duration. Nothing so philosophical as

an “equal right to happiness” is required contra Herbert Spencer (cf. J. S.

Mill, Utilitarianism, Ch. 5: par. 36, footnote).

(4) The u

’s are, as stated, numbers that measure the net balance of hap-

piness for each individual I over the relevant stretch of time (during which

the institution or action in question has effects). We can imagine this

stretch of time divided into unit intervals so each u

= u

,j= 1,...,q. But

this is just frills, so no more of this kind of thing. You can see how it goes.

(5) The fundamental point is that the u

’s represent only one kind of in-

formation: namely the net balance of utility computed, or estimated, solely

from the intensity and duration of the agreeable or disagreeable conscious-

ness of pleasure and pain, regardless of any of the objective relations of in-

dividuals to one another that are conditions of these experiences, or the

aims of the desires the satisfaction or dissatisfaction of which brings about

pleasure or pain. In themselves, the pleasures of vindictive cruelty count

equally with those of generosity and affection. As Bentham said: at the

margin, push-pin is as good as poetry (a unit of one = a unit of the other).

§3. Some Comments about Interpersonal

Comparisons of Utility (IP-Comparisons)

(1) In order to speak of maximizing a linear sum of utilities, we must

assume that it makes sense to add the pleasures and pains of each individ-

ual and that the units in which these are estimated are the same for differ-

ent individuals. The classical doctrine assumes full comparability of inter-

personal comparisons: (a) that levels of happiness are comparable, and (b)

in the same units. The Bentham-Edgeworth-Sidgwick line also assumes a

natural zero, a point of indifference between pleasures and pains. On these

matters, see Sidgwick’s Methods of Ethics, Bk. II, Ch. 2.

(2) The classical doctrine assumes that each individual can estimate and

compare their own levels of happiness on the basis of introspection and

memory: pleasures and pains are aspects directly known in experiences

judged to be agreeable or disagreeable.

[ 389 ]

Four Lectures on Henry Sidgwick

9. [See Sidgwick Lecture III of this volume for an extensive discussion of interpersonal

comparisons of utility. —Ed.]

EXAM COPY

(3) Assuming the same natural zero for all individuals, and (following

Edgeworth) the same just-noticeable difference in levels of happiness as the

common unit for all individuals, as well as supposing that all individuals can

rank consistently the differences between levels of happiness, it follows that

the interpersonal comparisons required are forthcoming, and without rely-

ing on choices involving chance and risk. (These assumptions are extremely

strong and seem implausible; but more on this later.)

§4. Some Features of the Principle of Utility as the

First Principle of a Rational Method of Ethics

(1) We are especially concerned with those features of classical utilitari-

anism that lead Sidgwick to think it overcomes the defects of intuitionism,

as shown in his discussion of justice (discussed in the previous lecture).

With this in mind, notice ﬁrst that utilitarianism is a single-principle con-

ception: a conﬂict of ﬁrst principles is impossible since there is only one

such principle. This is a gain over intuitionism.

(2) Moreover, Sidgwick believes the principle of utility is the conse-

quence of three self-evident (or seemingly self-evident) principles: (a) the

principle of equity (or fairness) (ME, pp. 379f ), which Sidgwick ﬁnds al-

ready formulated in essentials by Clarke (pp. 384f ) and Kant (pp. 385f );

(b) the Principle of Rational Self-Love (zero time preference) (p. 381); and

however, do not conﬂict but together yield the single principle of utility;

and so the criterion of self-evidence is satisﬁed without giving up the crite-

rion of having a guide in practice. (On this see Lecture I on Sidgwick

at B: 5.)

(3) Sidgwick contends that the principle of utility is fully rational in that

it is not limited or restricted by arbitrary exceptions or qualiﬁcations; it ap-

plies with full generality to all cases of practical reasoning; and the use of

secondary rules or “middle axioms” (ME, p. 350) is explained by the princi-

ple itself, thus meeting the criterion discussed in Lecture I, B: 5.

(4) Finally, the utilitarian principle harmonizes and systematizes com-

mon-sense judgments and adjusts them in a coherent and consistent way.

(See, for example, the discussion of ideal values in Bk. III, Ch. 14, and the

conclusion (pp. 406f ) that no principle but that of utility can organize these

[ 390 ]

appendix

EXAM COPY