On Knowing the 'Why': Particularism and Moral Theory

Author(s): Margaret Olivia Little

Source: The Hastings Center Report , Jul. - Aug., 2001, Vol. 31, No. 4 (Jul. - Aug.,
2001), pp. 32-40
Published by: The Hastings Center

Stable URL: https://www.jstor.org/stable/3527954

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/3527954?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

The Hastings Center is collaborating with JSTOR to digitize, preserve and extend access to The
Hastings Center Report

This content downloaded from

105.154.225.48 on Sat, 16 Mar 2024 03:30:02 +00:00
All use subject to https://about.jstor.org/terms
 On Knowing
the 'Why'

Particularism
and
by MARGARET OLIVIA LITTLE Moral Theory

If particularism is right, the broad moral claims we make are usually riddled with exceptions. But such

generalizations can still be a useful, even necessary part of moral life. They help us show what we should do, and

they are essential for understanding why we should do it.

oral particularism-or situationism, as it has Contemplation of such an idea often provokes

sometimes been called-seems to present an
especially radical objection to the enterprise
of moral theory. While the project of many "antithe-
orists" has been to battle philosophers' tendency to-
ward arid or overly tidy pictures of morality, particu-
larism seems to put pressure on the point or possibil-
ity of doing moral theory at all. It argues, (in)fa-
mously, that the moral import of any consideration is
irreducibly context dependent, that exceptions can be
found to any proffered principles, and that moral
wisdom consists in the ability to discern and interpret
the shape of situations one encounters, not the abili-
ty to subsume them under codified rules.' The posi-
tion thus seems, in the minds of many, to suggest not
that moral theory needs to be richer than has been its
makers' wont, but that there is no such thing, or at
least no such thing we need.
feelings of vertigo-not to mention derision. But it
also provokes confusion. Is the particularist really say-
ing that theory has no place in the moral life? Such a
view seems curious. After all, the philosopher most
often claimed as an ally to particularism, Aristotle,
didn't seem to eschew theory so entirely. He didn't
confine himself to commenting on individual cases;
and he insisted that the person of moral wisdom
must know the 'why,' not just the 'that'-something
that sounds, one might have thought, like a call to
theoretical abstraction. Add to this the fact that many
particularists agree that principles have some role or
other, and it's fair to wonder whether some notion of
theory is compatible with particularism-and what
that notion might be.
As a card-carrying moral particularist who makes
a living doing something I'd be happy to call moral
theory, I do think the two are compatible. But I think
the insights underlying particularism offer profound-

Margaret Olivia Little, "On Knowing the 'Why': Particularism and

ly important lessons on how moral theory should be
Moral Theory," Hastings Center Report 31, no. 4 (2001): 32-40. conceived. For one thing, theory turns out to be less

32 HASTINGS CENTER REPORT July-August 200 1

This content downloaded from

105.154.225.48 on Sat, 16 Mar 2024 03:30:02 +00:00
All use subject to https://about.jstor.org/terms
central to the moral life than certain
traditions have thought. More im-
portantly, particularism, properly un-
derstood, presents a different picture
of the kinds of generalizations that
make up moral theory-that make
up, in the end, our understanding of
the 'why.'
Varieties of Antitheory
want to start by isolating the dis-
tinctive challenge that particular-
ism seems to pose for moral theory.
As many have noted, the objections
presented under the "antitheory"
rubric form a diverse class.2 Some-
times, the point has been to object to
ambitiously reductive theories, in
which a very few concepts are said to
be able to generate all of our moral
considerations if only we spin them
out properly-such as Kant's theory
that all of morality (including the
virtues, as it turns out) could be gen-
erated out of the notions of respect
for self and others.3 Such views, it is
argued, buy simplicity at the cost of
accuracy: the moral landscape cannot
be understood by reference to some
one or two concepts.
Other objections target theories
that bleed out any distinctive role for
judgment. On some treatments of
morality (hedonic utilitarianism
comes to mind), the considerations
said to have moral import are ones we
could in principle design a sensor to
detect for us, and their relative
weights something a computer could
render algorithmically (they are lexi-
cally ordered, say, or commensurable
in the strong sense that renders all
weightings quantitative). In contrast,
it is argued, moral expertise just isn't
the sort of thing a machine could
have. It takes interpretation to deter-
mine when an action counts as mer-
ciful, and again when the demands of
mercy trump those of justice instead
of the other way around.
At other times, the objection is to
theories preoccupied with an overar-
ching ideal. Thus Stuart Hampshire
objects to approaches that try to lo-
cate all of morality in what is com-
July-August 200 1
105.154.225.48 on Sat, 16 Mar 2024 03:30:02 +00:00
mon to rational beings or human rather than by approximation to an
creatures as such; some of the most ideal, and that requires judgment in
important moral values and directives its application. In short, the "antithe-
in life, he urges, are those that flow ory" objections just canvassed are per-
from social practices or ways of life fectly consistent with the idea that we
that can permissibly vary from person can and should build a moral theory.
to person or culture to culture.4 Pur- Somewhat more radical are objec-
suing a related theme, Annette Baier tions to the idea that the moral ter-
criticizes the idea that we best gain di- rain forms any "unified system" at all.
rection on how to live by modeling a As classically used, the notion of a
morally ideal world.5 Given that our system is more than just an amalgam
own struggles are animated most cen- of all true propositions about some
trally by the presence of moral imper- subject matter, the idea of unity more
fection, it is folly to think we can best robust than mere completeness and
understand morality by starting with consistency. Rather, the phrase im-
a removed ideal-say, the Kingdom ports the notion of a well-ordered set,
of Ends-and then adding layers of in which a finite number of concepts
failure, rather than imagining incre- suffice to capture the terrain, in
mental improvements to the imper- which each concept is related to the
fections we confront. others in a codified, law-like way, in
Many "antitheory" objections are perfectly
consistent with the idea that we can and should
build a moral theory.
And of course, many have object- which axiom can be ordered to pos-
ed to proposals eliminating elements tulate, and which admits of a neat
that seem, on reflection, essential to taxonomy whose crisp edges are
the richness of moral life-to theories marred only by whatever vagueness is
that deny the existence of genuine inherent in its member concepts.8
dilemmas or the moral importance of In contrast, some have urged, the
emotion, or again, to theories that categories that together make up the
seem to think that merely possessing moral terrain are a motley crew. The
a moral theory is enough to make us concepts needed to capture the ter-
moral.6 rain are open-ended, and are often
I myself think all these objections orthogonal to each other or assume
are important (if sometimes levied a overlapping shapes that don't cleanly
bit indiscriminately in discussions of fit any genus-species taxonomy. Thus
historic figures7). But whatever one Amelie Rorty argues against attempts
thinks of them, they are clearly objec- to recover some final unity out of the
tions to impoverished moral theory, concepts-indispensible, every one
not to moral theory per se. Someone -belonging variously to virtue theo-
persuaded of the above criticisms ry, deontology, and utilitarianism. To
could nonetheless find a moral theo- do so would be to bleed out the sub-
ry to make her happy-one that is stantive historical allegiances that give
pluralist about values and duties and each approach its greatest wisdom.9
nonreductive about what carries the- And Iris Murdoch argues that the cat-
oretical weight, that admits of dilem- egories needed for accurate moral de-
mas and the importance of directives scriptions of the situations one con-
specific to idiosyncratic ways of life, fronts (the central task, she thinks, in
that addresses imperfection directly moral life) is a thoroughly open-
HASTINGS CENTER REPORT 33
This content downloaded from
All use subject to https://about.jstor.org/terms
 ended affair. There is no notion of an for an action, can in another counttips for rather than against the con-
endpoint to that discovery: moral de- against it or be irrelevant, and all in aclusion.
scription is unboundedly rich.1? On way that cannot be cashed out in fi- Moral particularists vary in how
these views, then, it's not that we nite or helpful terms. Pain is bad-
broadly they cast their claim of con-
need to acknowledge a richer system well, except when it's constitutive of
text-dependency. Some believe it is
in morality, it's that strictly speaking, athletic challenge; intentionally only so-called naturalistic features
morality forms no system at all. telling a falsehood is prima facie(those describable without obvious
This sort of objection is obviously wrong-well, but not when done to use of evaluative language) that carry
more radical than the first. Nonethe- Nazi guards, to whom the truth ismoral import holistically; moral con-
less, its acceptance is still consistent not owed, or when playing the game siderations so identified are granted
with an important notion of theory. Diplomacy. Pleasure always counts ininvariant reason-giving force-that
Even if morality doesn't admit of "a" favor of a situation-well, exceptan action is just always counts in its
theory, understood as a well-ordered when it's the sadist's delight in herfavor, that it causes pleasure does not.
model, we can still theorize about the
moral realm. After all, for all that's
been said, we can give definitions of
all those many useful concepts-out-
lining when cases fit under their um- In real life, moral views are changed by
brellas-or sort them as values and
disvalues, as things to pursue and experience and art as much as by argument:
things to avoid; and in doing so we
would have constructed bits of moral someone's sexist views about women change
theory. "Interference without con-
sent is a moral violation;" "generosity by fighting alongside them in battle, their
is the mean with respect to sharing
scarce resources;" "pain is a bad;" views of unfettered capitalism are undermined
"kindness is solicitous concern for an-
by working as a night janitor, and their views
other's well-being," and so on. We
may not be able to corral these pieces about eating animals shift by reading a poem.
of theory into one unified model, but
we can articulate the pieces them-
selves. What we can still have, in victim's agony, where her pleasure is For others, it's in for a penny, in for a
short, are theoretical generaliza- precisely what is wrong with the situ- pound: even 'cruelty' is said to switch
tions-broad claims that do more ation, not its "moral silver lining."1 valence depending on the context in
than assess individual cases, and do The claim is not just that the which it appears, and the aphorism
so, not just contingently (as when we moral contribution made by these that you sometimes have to be cruel
considerations gets outweighed by to be kind is to be taken at face
point out that Dora is usually cruel to
Jack), but as explanatory generaliza-others (as when the pain of a shot is value.12
justified by the utility it brings); the
tions that seem to get at the nature of And whether the valence of such
various moral considerations. These claim is that the moral "valence" of moral properties varies or not, some
are the constitutive elements of theo-the consideration, as it were, itself de- particularists will look skeptically at
ry. Moreover, one might well think, pends irreducibly on the context in the idea that we can succeed in pro-
they are the most important such ele-which it appears. The claim, in viding them with exhaustive, non-
ments, for they are what allow us to essence, is the cousin to claims of trivial definitions.13 Generosity is the
explain, to extend our knowledgeholism in the theory of knowledge. mean with respect to sharing scarce
most ambitiously, to make explicit Having a perception of a red cup can resources-well, except when the
the basis for criticizing others. provide excellent evidence that there generosity in question is the generos-
But it is here that the particular-is indeed such a cup; but there are ity of interpretation. Cruelty is wan-
ist's position seems to present a dis-contexts in which having such an ap- ton infliction of pain-well, except
tinctive challenge. For it is this verypearance counts precisely against when the cruelty is constitutive of
notion-the building blocks of theo-drawing that conclusion-as when kindness. Again, the scope of skepti-
ry, as it were-that particularism you know the evil demon is playing cism here need not be universal: vir-
seems to press on. The particularistwith your eyesight-and there's no tually all will agree we can recover ex-
argues that considerations carry theircashing out in finite or helpful terms ceptionless generalizations if we get
moral import only holistically. A con-the contexts in which the evidence sufficiently abstract. The question is
sideration that in one context counts whether those areas not admitting of

34 HASTINGS CENTER REPORT July-August 200 1

This content downloaded from

105.154.225.48 on Sat, 16 Mar 2024 03:30:02 +00:00
All use subject to https://about.jstor.org/terms
such generalizations are now placed license some inference; she is instead
it is to have mastery of a concept
beyond the bounds of theory. justified in believing the rose yellow(whether 'proton,' 'chair,' or 'cruelty')
The particularist's position thus because her faculty of sight is in good
is to possess the ability to see directly
seems to cast into doubt, not just working condition. So, too, we can rather than infer its instances in the
whether bits of theory can be well-or- world. And we can teach others, not
be justified in our moral conclusions,
dered in the classical sense, but not just by subsuming the situation just by defending arguments, but by
whether for large portions of morali- under some general law, but by tak- introducing, endorsing, and high-
ty, at least, we can isolate any bits of ing in the situation, exercising dis-
lighting some concept as more apt for
theory to lean on. cernment and wisdom-Aristotle's interpreting a given situation than
phronesis-and seeing that it's cruel. the one that might traditionally be
Questioning the Need for These are important points, and used.
Theory we can move from Aristotle to It is absolutely true that all of
Wittgenstein to explain the general these moves-teaching, criticism,
ow one of the points particular- justification-involve the need for
insight. To teach is to do something
ists rightly insist on when such that enables another to catch on toand invocation of generality. But a
claims raise a worried eyebrow is that the shape of a concept; to convert concept
or is, perforce, something of
the need for theoretical generaliza- dislodge a view is to get someone generality
to and abstraction. It applies
tions is badly overblown by theory- apply a different concept from the to many individuals; it groups them
loving philosophers. On a view famil-one they were deploying; and these together in virtue of what they share
in common. Thus when Julia Annas
iar from the Enlightenment, theoreti-are changes we can effect in all sorts
cal generalizations are at the veryof extra-theoretic ways. Certainly, we
argues that moral expertise requires a
"principled" understanding of moral-
heart of all sorts of moves integral to don't need to provide some set of nec-
the moral life. We're told that to per- essary and sufficient conditions: few
ity, everything depends on what is
suade rationally (as opposed to con- concepts even admit of such defini-
meant by "principled."17 If it's read as
verting by head blow), we must give tions. A chair is functionally under-
a requirement that one possess excep-
arguments; to teach, we must give de- stood as something to sit on-well,
tionless definitions or codified gener-
alizations, it simply isn't true. For
finitions; to be able to critically dis- except that many things that aren't
lodge our own or others' faulty intu-chairs (including Uncle Fred) can be
"principled" can mean mastery of the
itions on more than a case by casegrand to sit on, and many things that set of relevant concepts-having a
basis, we must appeal to broad princi-are chairs, such as ornamental chairs
deep understanding of the concepts,
ples; to justify our intuitions about made intentionally frail, are things we
not just surface competence, and the
cases, we need to subsume those intu-can't sit on at all.15 There is no saying
skill to navigate them when they tan-
itions under, or secure their coher- once and for all what counts-whengle together in concrete situations.
ence with, articulated generalizations.the object at the Museum of Modern Nor is it true, as it's often claimed,
But it's just wrong to think we can Art is a chair, or a work of art, or
that we are limited to dislodging
accomplish these important aims both. If a concept as easily learned opinion
as case by case if we don't have
only by appeal to explanatory gener-'chair' resists such attempts, wetheory. Where we succeed in intro-
alizations.14 In real life, we teach in all shouldn't be surprised that concepts
ducing 'subordination' as a telling
concept, or in shunning the appro-
sorts of ways-telling fables, pointingsuch as 'kindness' similarly resist cap-
to exemplars, interpreting a caseture. It takes experience-sometimes priateness of 'chastity,' we have suc-
jointly witnessed. In real life, morala lifetime-to understand the nuance
ceeded in one fell swoop in changing
others' opinions on an enormous
views are changed by experience and that separates tough love from aban-
art as much as by argument: some- donment. number of cases. Generality, in short,
one's sexist views about women We may not be able to spell out is not just found in theoretical gener-
change by fighting alongside them in alization.
any but the most trivial definition,
then, but this doesn't keep us from
battle, their views of unfettered capi-
Recovering a Role for Theory
being able to catch on, or get others
talism are undermined by working as
to catch on, to the shape and point of
a night janitor, and their views about
eating animals shift by reading a concept. (To be rule-guided in ll of this strikes me as clearly
Wittgenstein's sense one does not
poem. And while justification some- correct, so far as it goes. But
times proceeds by subsumption need access to codified principles,
how far is that? Are we really to think
under, or coherence with, explanato- from all of this that we never make
only a sense of how to go on.16) Once
we do catch on, we can come to see
ry generalizations, it doesn't always. use of theoretical generalizations in
Someone who sees that a rose is yel-that a situation is kind, or cruel-not
morality? Aristotle-the great advo-
low, for instance, is hardly appealing cate of phronesis-didn't think so.
because we have some spooky faculty
of
(even tacitly) to a general principle to moral intuition, but because His
what treatise on the virtues is suffused

July-August 200 1 HASTINGS CENTER REPORT 35

This content downloaded from

105.154.225.48 on Sat, 16 Mar 2024 03:30:02 +00:00
All use subject to https://about.jstor.org/terms
 with theoretic moves. He didn't just
point to exemplars or tell stories; and
while he insisted that moral wisdom
requires life experience, he saw fit to
articulate all manner of theoretical
generalizations intended to guide and
illuminate what that experience will
reveal.
And so do we. If we reflect on our
shared moral life, it certainly looks as
though an important part of how we
justify, convince, teach, and clarify is
by pointing to explanatory general-
izations whose truth we seem to en-
dorse. Sometimes we convert by
showing a film; then again, some-
times we do it by giving an argument
(say, that one shouldn't discriminate
on the basis of sex). Sometimes we
teach by modeling behavior; but
sometimes we do it by articulating a
generalization (say, that wrongful in-
terference is measured by lack of con-
sent). And when we want to under-
stand what someone means when she
invokes a contested concept (say,
'equality'), sometimes we ask for her
verdict on a test case, but sometimes
(if only to control for differing factu-
al interpretations) we ask her to give
us her definition. In short, we seem
to theorize-to appeal to explanatory
generalizations-about morality all
over the place.
Particularists often respond to this
point by re-interpreting the service to
which these generalizations are put.
Such principles, it is said, are useful
pedagogic devices-helpful crutches
for novices in moral judgment who,
like beginning cooks, often need in-
tentionally simplified rules or recipes
to guide them as they gather needed
experience. Or again, they are valu-
able heuristics: they aid our ability to
interpret a case by serving to remind
us of what can be salient. Or again,
they are summaries of past cases, use-
ful not just as records of history, but
as shorthand memos of what tends to
be relevant. Or again, they are rules
of thumb, giving us a set of presump-
tions ("don't stab!") to take with us
when we head out into the world.
These are surely some of the uses
to which moral principles are put.
36 HASTINGS CENTER REPORT
105.154.225.48 on Sat, 16 Mar 2024 03:30:02 +00:00
All use subject to https://about.jstor.org/terms
But to think these functions exhaust
their use seems profoundly at odds
with moral practice. When we teach
our children that it is morally prob-
lematic to inflict pain on their tod-
dler siblings (to give an example dear
to my heart just now), we're surely
not just saying that infliction of pain
can have moral import. That's some-
thing we can say of anything, includ-
ing shoelace color. But the morally
wise person, one would have
thought, is someone who under-
stands that there is a deep difference
in moral status between infliction of
pain and shoelace color, even if both
can, against the right narrative, be
bad-making.
Nor is it enough to say that moral
generalizations just assert what usual-
ly carries moral import. To be sure,
many of the moral presumptions we
arm ourselves with are inductively
based, as it were: part of why I pre-
sume not to lie to those I meet is be-
cause, fortunately, my misanthropic
enemies are few and far between.
This, though, is a thoroughly contin-
gent fact. There are possible worlds
(life in your favorite post-apocalyptic
movie) in which enemies outnumber
friends by a wide margin, and most
of the lies in fact told are honorable
rather than shameful. Are we really to
think that the generalizations we
work so hard to isolate are just meant
to be statements of local frequen-
cies-as though the only thing we
mean when we make generalizations
about the morally problematic nature
of lying is that, in our neck of the
woods, the situations in which lying
is wrong-making outnumber those in
which it is not?
So, too, with principles' claimed
pedagogic function. Sometimes the
directives we issue are very crude ones
(the sweeping "never lie!" uttered to a
five-year-old), meant to be left be-
hind once one moves from novice to
expert. But it is difficult to imagine
that the hard-won insights of philos-
ophy discussions are all just tempo-
rary crutches toward enlightenment;
they sometimes seem, not what the
person of practical wisdom leaves be-
This content downloaded from
hind, but precisely what she under-
stands.18
But if particularists are often defla-
tionary in their views of moral gener-
alities, it's because it can seem unclear
how they can say anything more ro-
bust. Exceptions, after all, are every-
where: anything could have, and
nothing must have, moral import.
Such puzzlement can lead to skepti-
cism about the very possibility of
doing theory beyond endorsing or
objecting to the use of various con-
cepts: one may conclude that there is
no room left for theoretical moral
generalizations.
I think particularism carries a dif-
ferent lesson. What it really invites us
to consider are statements that are
law-like despite admitting irreducibly
of exceptions. What it really invites
us to think about is what Aristotle
might have meant by the claims he
called "for the most part" generaliza-
tions.
Defeasible Generalizations and
Moral Theory
he qualifier "for the most part"
gets bandied about rather casual-
ly, as do its sometimes substitutes "ce-
teris paribus" and "all things equal."
These phrases are often used inter-
changeably as all-purpose qualifiers
whose meaning is left opaque. But
what do we mean when we advance a
generalization thusly qualified? If we
say "ceteris paribus, lying is wrong-
making," or "for the most part, pain
has a negative valence," or "the
prospect of pleasure counts, all things
equal, in favor of an action," just
what is it that we are saying?
Some read these claims as purely
statistical ones (understandably, real-
ly, given the quantitative ring of Aris-
totle's phrase). On this interpretation,
use of "for the most part" means,
quite literally, that the asserted con-
nection between, say, lying and
wrong-making, or pleasure and good-
making, holds with high frequency.
To read such generalizations in this
way, however, is to give up any pre-
tense that they are robustly explana-
July-August 200 1
tory. Except in areas like quantum
mechanics, which are ruled by gen-
uinely statistical laws, statistical gen-
eralizations are contingent ones. (If
catfish "usually" weigh about four
pounds, this means not just that peo-
ple occasionally catch six-pounders,
but that there are some rivers-and
possible worlds-in which most of
them tip the scales at six.) On this in-
terpretation, then, the connection as-
serted doesn't get at the nature of the
subject it concerns, but at the fea-
tures the subject usually displays as a
matter of fact.
Those explicit in maintaining the
explanatory nature of "for the most
part" generalizations, on the other
hand, usually do so by interpreting
that qualification as a signal that
we're talking shorthand. The asser-
tion, it turns out, is an "en-
thymeme"-a claim containing sup-
pressed premises we could fill in if we
just had the time. (There is a con-
crete exceptionless generalization in
the offing, we just don't quite know
yet what fills in the gaps-or it's so
obvious we needn't bother to state it.)
On this reading, "for the most part"
generalizations are indeed explanato-
ry, but they are thought capable of
serving that function only because
the exceptions are in principle elim-
inable. Here, explanation is a species
of deduction.
On the first, statistical, interpreta-
tion, qualified generalizations are not
robustly explanatory-they are, at
best, locally useful as inductive
guides to one's neighborhood. On
the second interpretation, qualified
generalizations are explanatory, but
only because the qualification is ca-
pable of being expunged.
These interpretations seem to me
to get exactly wrong the sort of gen-
eralization Aristotle was actually
hinting at.19 While one could think
these options exhaust the possibili-
ties, it is also possible-and truer to
Aristotle's own views-that the les-
son here is a quite different one:
namely, there are generalizations that
are porous and genuinely explanato-
ry.20
July-August 2001
105.154.225.48 on Sat, 16 Mar 2024 03:30:02 +00:00
When we issue a generalization to lead, to be sure; but when they do,
the effect that something has a cer- it's a sign that something is epistemi-
tain feature, sometimes what we real- cally amiss. Someone who has en-
ly want to say is not that such a con- tered the circus's Hall of Holograms
nection always, or even usually, has entered a situation that, however
holds, but that the conditions in fun, is defective. If the "evidential va-
which it does hold are particularly re- lence," of appearances turns negative,
vealing of that item's nature. We it's a signal that one has entered a de-
might put it by saying that we're as- viant context. Appearances, we
serting what happens in "normal" might put it, have a default evidential
conditions, except that the notion of valence of being trustworthy; put in
'normalcy' is so freighted with mis- another way, they are defeasibly trust-
leading connotations. Better put, worthy.
then, we are taking as privileged, in When we say "all things equal, ap-
one way or another, cases in which pearances are the sorts of things we
the item has the feature specified. can take at their word," then, we are
There are various ways in which that not claiming that appearances are
privileging move can take place; I'll usually trustworthy. Someone who
mention two prominent ones. In gets stuck in the Hall of Holograms
one, the conditions are elevated in an may never again be able to trust her
evaluative sense: where the connec- eyes; the unfortunate brain in a vat is
It is difficult to imagine that the hard-won insights
of philosophy discussions are all just temporary
crutches toward enlightenment; they sometimes
seem, not what the person of practical wisdom
leaves behind, but precisely what she understands.
tion fails to hold, it means that some- misled most of the time. What we are
thing has gone awry. In another, we saying, rather, is that such a situation
are saying that cases instantiating the is thereby marked as defective by
feature in question count as a para- knowledge's own lights. The situa-
digm against which departures are tion is deviant, not in a statistical,
understood. but in an evaluative sense.
For an example of the first, return In the second sort of case, we
to the case we used in illustration of mark exceptions as deviant in a dif-
holism in the theory of knowledge. ferent way. To illustrate, return to our
While having a perception of a red example of the concept 'chair.' Orna-
cup often counts as an excellent rea- mental chairs, we noted, are still
son to think such a cup sits beforechairs even though we can't sit on
them. Nonetheless, it seems natural
you, we noted, there are all sorts of
to think there is some sort of inti-
cases in which it counts in just the
opposite way, as when you remember mate connection between 'chair' and
you've taken an hallucinogenic drug. the function of holding people in re-
Nonetheless, it seems natural to pose; and we might intuitively think
think there is some sort of intimate to put the point by saying something
connection between appearances and like "ceteris paribus, chairs are things
justification. Appearances can mis- we can sit on." Once again, such a
HASTINGS CENTER REPORT 37
This content downloaded from
All use subject to https://about.jstor.org/terms
 claim is not a statistical one. We're might call defeasible generalizations:
a bad-making feature of the situation
not committed to saying we can usu- generalizations that privilege the con- that lying is now a moral plus: would
ally sit on chairs: a very opulent soci- ditions or cases in which a certain that it weren't honorable here to lie.)
ety might in fact have more orna- connection holds. This is also, I be- Telling a lie while playing Diplomacy,
mental than functional chairs lying lieve, just what Aristotle was doing in turn, involves nothing morally
about. But we need not be finding (or often doing) when he offered his amiss, but understanding its moral
fault here with ornamental chairs. It's theoretical generalizations. When he status relies on invoking a notion,
not as though there is something de- talks about the "usual" nature of re- consent, that itself cannot be under-
ficient or defective about them paying one's debts, say, he is making stood without invoking a framework
explicit its default status, and, in in which the normal case is not to lie.
(they're no good to sit on, to be sure,
but that doesn't keep them fromdoing so, he is implicitly calling our And while pain can be an integral
being fabulous-and fabulouslyattention to what is theme and varia- and even cherished part of the athlet-
sought-after-chairs). What wetion, deviance and normality, para- ic challenge, we understand it as pain
mean, instead, is that there is some-digm and emendation. only if we understand that this situa-
thing like theme and variation in- On this theory, when we say "ce- tion is the riff, not the theme. And if
volved. The ornamental chair is, if
teris paribus, lying has a negative va- generosity cannot be defined as the
you like, a riff on the theme of chair;
and one can't understand a riff with-
out understanding the theme to
which it stands as variation. The priv-
ileging move here, then, is not about
what is better, but about what has, as What theory turns out to be necessary for is
it were, conceptual priority: to under-
understanding-for knowing what Aristotle
stand something as an ornamental
chair one must understand the no-
called the 'why' rather than the 'that.'
tion of chairs that are for sitting on,
but not vice versa.
In neither of these cases, then, are
we saying that the items in questionlence," we are not saying that it al- mean with respect to sharing scarce
usually have the features we high- ways, must, or even usually has that resources, we may still think its na-
light. Nor, though, must we think westatus; we are saying that this is the ture is best illuminated by regarding
can exhaustively specify the condi- valence it has in conditions that are that notion-or even the metaphors
tions under which they in fact privileged in various ways. When we it evokes-as paradigmatic.
would-filling in just when a chair say "all things equal, pain is bad-mak- Thus it is not true that particular-
ing," we are not saying that pain al- ism allows no theoretical generaliza-
can and cannot be sat upon, or speci-
ways carries this valence, nor merely tions. It can acknowledge defeasible
fying once and for all the conditions
in which appearances are and are notasserting that it usually does in our generalizations.22 These are complete-
neck of the woods. We are saying, in-
trustworthy. For one needn't think we ly consistent with even a radically
can specify the conditions in which stead,
a that pain is defeasibly bad- holistic doctrine of morality, accord-
connection does obtain in order to making; it has a default negative va- ing to which we cannot identify once
say that where it does counts as a priv- lence.21 Where lying and pain lack and for all, in any concrete terms, the
ileged case. Such generalizations tell this valence, as they sometimes or contexts in which a given considera-
us about the nature of something, in even often do, it is because they occu- tion counts for or against an action.
essence, not by getting rid of excep- py a context defective by morality's For once again, one needn't be able to
tions to the isolated connection, but own lights, or again because the cases delineate the conditions under which
by maintaining and demarcating in question are operating as variations a consideration has a given import in
their status as exceptions. ("Excep- that cannot be understood except by order to elevate as privileged those
tions," again, that can outnumber the reference to a paradigm that carries cases in which it does.
rule in all sorts of contexts, for the the privileged valence. This means, to be sure, that using
measure of exception here is not a sta- Thus telling a falsehood to the such generalizations won't give us a
tistical one.) Nazi guard, for instance, may indeed moral analysis of lying, or interfer-
In short, "ceteris paribus," or "for be honorable rather than shameful, ence, or inflicting pain. We won't get
the most part" generalizations need but it's because something has gone from them some set of conditions
be neither statistical nor en- awry: there is something badly amiss laying out when lying is justified and
(namely, the Nazi's evil) from the unjustified, or even when it counts as
thymematic. When we use them, we
may instead be asserting what we
moral point of view. (It is, if you will, a moral plus instead of minus. The

38 HASTINGS CENTER REPORT July-August 200 1

This content downloaded from

105.154.225.48 on Sat, 16 Mar 2024 03:30:02 +00:00
All use subject to https://about.jstor.org/terms
particularist still insists that navigat- or the vulnerable agent, in our reflec- mastery doesn't involve appreciation
ing the moral world is at bottom a tions? How we resolve all of this is, of of defeasible generalizations. An inte-
matter of skill, including now a skill course, a messy matter, and not one gral aspect of knowing "how to go
at understanding and recognizing we can codify (the particularist gives on" is knowing what counts as, and
what is deviant and normal, what a model of resolving disagreement no having the skill to navigate, what is
paradigmatic and emendation, what more than she gives one of lying). deviant and normal, paradigm and
conceptually prior or central. Defea- emendation, theme and variation. If
But the point is that the particularist
sible generalizations, that is, will not is not, as the usual image has it, con- it is wrong to interpret Annas's no-
concretely specify what the moral na- fined to the sidelines of theoretical tion of principled understanding as
ture of something is. moral debate-as though all she can possession of codifiable generaliza-
What defeasible generalizations do do is watch the play from the bench, tions, it's equally wrong to think it
allow the particularist to do, though, at most throwing in the occasional empty of generalization altogether.
is to preserve a distinction between story or concept while others trade Even the particularist thus must
actions or dispositions that can, and theoretical claims. For better oracknowledge an indispensable role
those that cannot, properly be said to worse, particularists can-and do-for the explanatory generalizations
have a moral nature. Shoelace color join the fray. that make up the theoretical. To agree
and infliction of pain can both be to this, though, is a far cry from
bad-making. In acknowledging this, agreeing to the goal of theory as tra-
though, we needn't be committed to ditionally construed. The particular-
the radical particularist thesis that f particularism is right, most of theist's lesson about the nature of ex-
they are therefore on a par (as moral generalizations we deploy inplanatory generalization counsels
though, if we single out the latter everyday life turn out to be irre-against the usual quest for theory,
rather than the former in our teach- ducibly porous. They are shotwhich is to spend all of our time fill-
ing and reflection, it is only becausethrough with exceptions we cannot ing in the holes of our generaliza-
doing so is locally useful). Shoelace eliminate. By the particularist's owntions. We get moral wisdom, in the
color doesn't have a moral nature; lights, though, these generalizationsend, not by filling in the exceptions,
lying does. The former doesn't have a can nonetheless count as robustly ex-but by knowing what counts as one
moral nature because, while it can planatory and insightful. Adducing in the first place.
have various moral imports (good- them has a power a list of instances
making, bad-making, indifferent) in does not, for it situates instances
various contexts, it has none of them within a framework that maintains Acknowledgments
defeasibly. It has, we might put it, no some as exceptions to others' rule.
Many thanks for helpful discussion
privileged import. Once we recover this sort of gen-with Alisa Carse, Jonathan Dancy, Jim
The availability of defeasible gen- eralization, it turns out that theory isNelson, Nancy Sherman, Gopol
eralizations also means that particu- in fact essential to moral life. Defeasi-
Sreenivasan, Jim Taggert, and members
larists can do more than endorse or ble generalizations will, as we noted,of my graduate seminar on moral par-
descry the use of various concepts be useful when it comes time to ticularism. Special blessings to Sarah
when trying to make progress in illu- teach, convert, and justify; but theseBuss, Hilde Nelson, Jim Nelson, and
minating the moral landscape. With- tasks, as we also noted, are ones thatMark Lance for helpful comments on a
out trying to exhaustively define the can in principle be effected throughprior draft.
paradigmatic case, or specify the con- moral discernment. What theory
ditions in which a privileged connec- turns out to be necessary for is under-References
tion holds, she can make explicit standing-for knowing what Aristo- 1. John McDowell is cited by many as in-
which defeasible generalizations she tle called the 'why' rather than the spiration for the view, which is developed
advances as true. Which types of ac- 'that.' For mastery of moral conceptsand endorsed more explicitly by, among
tions does she believe have a moral is mastery of defeasible generaliza- others, J. Dancy, Moral Reasons (Oxford:
Blackwell Publishers, 1993); D. Mc-
nature, in the sense used above? tions. To be sure, one can understand
Naughton, chapter 13 of Moral Vision (Ox-
What valence forms their default? the concepts without ever statingford: Blackwell Publishers, 1988); M. Little,
And, crucially, which sort of privileg- these generalizations explicitly (and "Moral Generalities Revisited," in Moral
ing-conceptual priority, centrality, one certainly need have no thought, Particularism (Oxford: Clarendon Press,
evaluative privilege-is she here de- as such, of the terms in which they 2000), ed. B. Hooker and M.O. Little. For
recent debate on the view as such, see B.
fending? This, in turn, is all fodder have here been put). But one cannotHooker and M.O. Little, eds., Moral Partic-
for debate. Does lying have a moral be said to understand moral conceptsularism.
nature in the sense used above? Do without appreciating the privileging 2. R.B. Louden, Morality and Moral The-
we illuminate 'respect for persons' moves that lie at their heart. Indeed, Iory: A Reappraisal and Reaffirmation (Ox-
best by centering the self-sufficient, doubt there are any concepts whoseford: New York, 1992). For a collection of

July-August 200 1 HASTINGS CENTER REPORT 39

This content downloaded from

105.154.225.48 on Sat, 16 Mar 2024 03:30:02 +00:00
All use subject to https://about.jstor.org/terms
 related essays, see S. Clarke and E. Simp- and Particularism," forthcoming in Notizie in a book tentatively titled "Porous Gener-
son, eds., Anti-Theory in Ethics and Moral di Politeia. alizations and Privileged Conditions." This
Conservatism (Albany: State University of 11. The pain example is from Milgram, section also signals an important change in
New York, 1989). the Diplomacy example from Mark Lance, the view I presented in "Moral Generalities
3. N. Sherman, Making a Necessity of and the pleasure example from Dancy, who Reconsidered," where I defended a more
Virtue: Aristotle and Kant on Virtue (New further attributes it to Roy Hattersley. radical particularism that denied moral im-
York: Cambridge University Press, 1997). ports to features such as lying-the import
12. A good representative of the first can
we precisely work here to recover.
4. S. Hampshire, "Morality and Con- be found in D. McNaughton and P Rawl-
flict," in his Morality and Conflict (Oxford: ings, "Unprincipled Ethics," in Moral Par- 21. Not an epistemic default, in the sense
Blackwell Publishers, 1983). ticularism, ed. Little and Hooker, pp. 256- of something we can justifiably presume is
75. An example of the second is in J. true until we see evidence to the contrary: a
5. A. Baier, "Theory and Reflective Prac-
Dancy, "In Defense of Thick Concepts," court servant at Versailles might wisely
tices," in her Postures of the Mind: Essays in
Midwest Studies in Philosophy, vol. 20 (Min- abandon his usual presumption of sitting
Mind and Morals (Minneapolis: University
of Minnesota Press, 1985). neapolis: University of Minnesota, 1995). on the chairs he encounters; the person who
realizes she is going into the hologram
6. To the objections just presented we 13. Note that the two sorts of objection
are different. One could be a holist about chamber rightly presumes to distrust what
can add two others that concern epistemo-
her eyes seem to tell her.
logical issues less relevant to the discussion the reason-giving force of considerations
22. These are different and more robust
pursued above. When Bernard Williams and yet believe moral concepts capable of
nontrivial exhaustive definitions. Note, too, than the explanatory generalizations I de-
objects to theory, for instance, his central
fended in the last section of "Moral Gener-
point is to criticize the idea that it might that one may well believe there remain
pockets of well-ordered theory even if the alities Reconsidered." There my concern
admit justification that is transcendent of a
culture's norms. See his Ethics and the Lim- whole is not so systematic. (Many, for in- was to provide an analysis of the "good-
stance, will believe the cluster of concepts making" relationship (not the notion of
its of Philosophy (Cambridge, Mass.: Har-
dealing with interference, agency, and con- moral nature here discussed); the explana-
vard University Press, 1985). And some-
sent form such an internally codifiable sys- tions recovered there hence apply equally
times "antitheory" is used to describe objec-
tem, however messily it might interface well to something like shoelace color, wher-
tions to the claim, no longer current, that
with other moral considerations.) I myself ever it does in fact count as a good-making
justification always proceeds in a "top- feature.
down" fashion-that is, that intuitions think it motley all the way down; but again,
about generalizations always carry more this is only an issue of how widely the par-
ticularist wants to cast her net.
weight than intuitions about individual
cases. 14. This is the central point of Cora Di-
amond's, "Anything
7. Kant, for instance, is often singled out But Argument?" The
reference
as a particularly egregious sinner in thesetore-
animals and poetry in this para-
graph
gards, but as many have pointed is a reference
out, he to her "Eating Meat and
Eatingor
never denied the need for judgment People,"
dis-both from her The Realistic
Spirit:
missed entirely the importance of Wittgenstein, Philosophy, and the
emotion
Mind (Cambridge, Mass.: Massachusetts
in the human moral life. And Aristotle,
often touted as the exemplar forInstitute of Technology Press, 1991).
antitheo-
15. My thanks
rists, is often let off the hook too lightly. As to Mark Lance for the ex-
ample. was
Hampshire notes (see ref. 4), Aristotle
dismissive of the idea that moral values and
16. See M.O. Little, "Wittgensteinian
directives located in "conventions," such
Lessons as
on Particularism," in Wittgenstein-
the way society organizes familialianor sexual
Bioethics, ed. C. Elliott (Durham, N.C.:
relations, could nonetheless be fully
Dukeauthor-
University Press, 2001), and J.
itative for those within those ways of "Particularity
Garfield, life. and Principle: The
And while not a reductivist, he was a true-
Structure of Moral Knowledge," in Moral
born systematist of the sort objected to in
Particularism, ed. Hooker and Little, pp.
the following paragraphs (although
178-204. he
agreed that the concepts needed to 17.
describe
J. Annas, "Virtue as a Skill," Interna-
morality are open-ended): each virtue is as-
tional Journal of Philosophical Studies 3, no.
signed its domain, and all are capable of
2 (1995): 227-43.
being corralled under one notion-flour-
18. See ref. 10., Milgram, "Murdoch,
ishing-to boot.
Practical Reasoning, and Particularism."
8. Most everyone, of course, will agree
19. For an argument defending the claim
that moral concepts can all be organized that Aristotle himself could not have en-
under very abstract concepts, though dis-
dorsed a statistical interpretation of "for the
agreement will persist in the extent to
most part" generalizations, but which then
which those abstract concepts thus do any
real work for us. tries therefore to render them in principle
exceptionless, see T.H. Irwin, "Ethics as an
9. A. Rorty, "The Advantages of Moral Inexact Science: Aristotle's Ambitions for
Diversity," Social Philosophy and Policy 9, Moral Theory," in Moral Particularism, ed.
no. 2 (1992): 38-62.
Hooker and Little, pp. 100-129.
10. I. Murdoch, The Sovereignty of Good
20. In this section, I am presenting work
(London: Routledge Kegan Paul, 1970); E.
done jointly with my colleague Mark
Milgram, "Murdoch, Practical Reasoning, Lance, which we are developing more fully

40 HASTINGS CENTER REPORT J u ly-August 200 1

This content downloaded from

105.154.225.48 on Sat, 16 Mar 2024 03:30:02 +00:00
All use subject to https://about.jstor.org/terms