Journal of Medicine and Philosophy 0360-5310/99/2404-0365$15,00

1999, Vol, 24, No, 4, pp, 365-375 © Swets & Zeitlinger

Coming Home to Hume: A Sociobiological Foundation

for a Concept of 'Health' and Morality^

Kenneth F. Schaffner
George Washington University

ABSTRACT

Assessing the normative status of concepts of health and disease involves one in questions
regarding the relationship between fact and value. Some have argued that Christopher
Boorse's conception of health and disease lacks such a valuational element because it
cannot account for types of harms which, while disvalued, do not have evolutionarily
dysfunctional consequences, I take Boorse's account and incorporate some Humean-like
sociobiological assumptions in order to respond to this challenge. The possession of
moral sentiments, I argue, offers an evolutionary advantage (thus falling within Boorse's
definition of normal functional abilities). However, this does not amount to emotivism: on
the contrary, these sentiments can be the basis of a value system. This value structure
introduces the concept of sympathizing with a fellow being's suffering as the basis of a
normative dimension to disease. For example, it holds the disvalue of disease to lie in the
fact that disease involves suffering and functional limitations,The naturalistic Humean
type of account presented here thus jumps the normative-descriptive divide. When Boorse's
account is extended to include social sentiments and behaviors, a conception of health
emerges which is broader than Boorse's or Kass's, but narrower than the WHO's,

Key words: Christopher Boorse, disease, fact-value gap, health, David Hume, moral
sentiments.

All ethical inquiry at some point needs to provide an account of moral

value. Historically, philosophers have taken dramatically divergent ap-
proaches to grounding ethical norms, among them Kantian rationalism,
utilitarian descriptivism, and emotivist (dismissive) naturalism.^ It oc-
curred to me a number of years ago (and to others as well — see Engelhar-
dt, 1996, pp. 197-199) that one route to examining the relation between
fact and value might be via an examination of the concepts of 'health' and
'disease.' Though this topic has been an interestingly contentious one in
the philosophy of medicine, generating strong descriptivist views (e.g..

Correspondence: Kenneth F, Schaffner, M,D,, Ph.D,, University Professor of Medical

Humanities, 709 C Gelman, George Washington University, 2130 H, Street, N,W,, Wash-
ington, DC 20052, USA,
366 KENNETH F. SCHAFFNER

Boorse, 1977; 1997) and complex normative accounts (e.g., Engelhardt,

1996), it is possible that an attempt to grasp the concept of health at its
roots might provide a useful alternative perspective to the fact-value dis-
tinction. The thesis of this paper is that if we hegin from Boorse, and make
some plausihle Humean-like sociohiological assumptions and accept some
contemporary critiques of a sharp fact-value distinction, we can arrive at a
normative account of health and disease that outflanks some objections to
Boorse's original project.
The importance of being (reasonably) clear about the concept of 'health'
(and relatedly the disease concept) can scarcely be overemphasized. The
concept of health is arguably part of the foundation of medicine. Consider
in this regard what two physician-philosophers. Otto Guttentag and Ed-
mund Pellegrino, have contended is the essence of medicine. In their view,
this essence is characterized by one distinct end — the affirmative re-
sponse of the physician to a plea from the individual patient to help in
restoring or preserving that patient's health. (For those who dislike "es-
sences," the Guttentag-Pellegrino point of view can be seen as a broad
principle.) Kindred views to the Guttentag-Pellegrino view can be found
in the Hastings Center project on the 'The goals of medicine' — see their
1996 Report [hereafter GMR, 1996], p. S9. This view is also used as a
unifying theme in Schaffner and Engelhardt's recent Encyclopedia of Phi-
losophy article on "Philosophy of Medicine." From this Guttentag-Pel-
legrino view of the essence of medicine, it follows that any deeper inquiry
about medicine or the goals of medicine will have to understand the nature
of the key concept in that characterization: "health."
As already suggested, accounts of the concept of health have led to the
deeply perplexing and often contentious debate on this notion that has
taken place in philosophy of medicine over the past twenty-plus years. To
the first approximation, the debate is largely between the descriptivists
(represented by Boorse, 1977; 1997 and, probably, Daniels, 1985) and the
normativists (represented by Bngelhardt, 1996; Fulford, 1989; Nordenfelt,
1987; Reznek, 1987).^ I am going to use the Boorse definition, both as a
basis for discussion, and as a foil to extend the definition, so it is worth
reviewing exactly how Boorse characterizes the notions of health and
disease. In his 1977 article he writes:

1. The reference class is a natural class of organisms of uniform func-

tional design; specifically an age group of a sex of a species.
2. A normal function of a part or a process within members of the
reference class is a statistically typical contribution by it to their
individual survival and reproduction...
 COMING HOME TO HUME 367

3. A disease is a type of internal state which is either an impairment of

normal functional ability, i.e., a reduction of one or more functional
abilities below typical efficiency, or a limitation on functional abili-
ty caused by environmental agents.
4. Health is the absence of disease (Boorse, 1977, pp. 562, 567).

This statistical typicality definition has a certain elegance and plausibility,

and does make sense of many states and conditions that medicine charac-
terizes as diseases, but there are some problems with this simple defini-
tion.
One argument for why a pure (or at least a pure and simple) descrip-
tivism will not suffice can be found in one widely discussed objection
to Boorse's analysis (and raised, in point of fact, by Boorse himself) that
is known as the "universal disease" objection. This objection is that
common states and conditions that are typically characterized as "diseas-
es" do not conform to the Boorse definition. These common conditions
include as examples dental caries, atherosclerosis in individuals after a
certain age, presbyopia in the 50+ group (the "can't see" problem),
benign prostatic hypertrophy in men past 70 (the "can't pee" problem),
and postmenopausal osteoporosis in women (the "better not fall" prob-
lem). Boorse attempts to sidestep some aspects of the universal disease
objection, which he terms an "anomaly," by citing "environmental"
factors that can interfere with normal functioning, but he admits this
only goes so far, and ultimately this universal disease objection is left
as an "anomaly" requiring further work. There are other arguments against
Boorse's definition, including criticisms of his concept of function (see
Reznek, 1987) and his ignoring of inclusive fitness (Engelhardt, 1996,
pp. 199-203) that I will not have the space to discuss. Suffice it to say
that a number of commentators have argued that what is missing in
Boorse's account is some valuational element — for example that a
condition is a disease because it is undesirable — because it entails
suffering or some type of harm, even though that suffering may not have
evoiutionarily dysfunctional consequences though it is disvalued (recall
the can't see/can't pee/better not fall examples)."* But if a valuational
element is to be introduced, it would be most useful for clarificationai
purposes, to know on what kind of theory of value that addition depends.
Thus any prescriptive account, and there are many, owes us a theory of
value, or at least a sketch of that theory. (On this point compare Fulford,
1989, who agrees with this need for a theory of value, but does not
develop such a theory, and Reznek, 1987, who advances the sketch of
an account that values those things that lead to "good lives.") What I
368 KENNETH F. SCHAFFNER

intend to do in the remaining part of my paper is to consider a sketch

of a theory of value, and I will then look at the implications that this
simple, and it turns out biologically based, theory of value has for our
concept of health and the goals of medicine.
In order to avoid additional complications related to diseases that have
prima facie volitional and mental components (e.g., alcoholism), this pa-
per only considers some biological/traditionally medical diseases, and asks
what a "minimalist" value theory might look like, and be based on, that
would account for characterizing something as a disease.^ I want to argue
that one can actually best proceed from the biological Boorsean perspec-
tive, and employ a notion of a disease as a deviation from a statistical
norm, one that leads to increased morbidity (including decreased repro-
ducibility) or mortality, but I want to add another element that introduces
the missing prescriptive component.
To obtain the missing and needed prescriptive element, I add to the
Boorse definition a common minimal core extractable from three some-
what contrasting biological-foundational proposals made in the recent eth-
ics literature. This position has important affinities to one introduced by
David Hume centuries ago, though it is one that has largely been ignored
both in the literature on ethical theory in general and in bioethics in partic-
ular, though there are some signs that this may be in the process of change.
Hume's account is complex and some commentators have felt that it has
so many aspects that it borders on incoherence (see Selby-Bigge's com-
ments as quoted by Garrett 1997, 187-188, but also Garrett's defense of
Hume in his chapter 9). The relations among reason and emotions or
sentiments in Hume are multifaceted. Though morals appear dependent on
empathy and sympathy with one's fellow beings, socialization within the
family plays a significant role. Reason both can correct sentiment (com-
pare Garrett, pp. 193-195), as well as be integral to tbe matter of moral
justice (see Hume, THN, pp. 484-491).^
First, we might note as Hume wrote "To the most careless observer,
there appear to be such dispositions as benevolence and generosity; such
affections as love, friendship, compassion and gratitude" that are "distin-
guished from those of the selfish passions..." Tenderness toward off-
spring is especially remarked on. From such a perspective, it is natural to
want to prevent illness and assist ill individuals to recover, and especially
to want to help immediate family. But Hume suggests that these senti-
ments can and do generalize, though they may not be as forcefully felt for
those at a distant remove. Nonetheless, more complex sentiments of moral
blame do hold sway in such general contexts, and they are also buttressed
by the conventions of justice.
 COMING HOME TO HUME 369

And Hume's ethical theory is, as Beauchamp has argued, best compre-
hended by considering Hume's moral philosophy as based on three inter-
connected propositions:

1. A moral psychology of reason and sentiment, in which desires and

sentiments determine our interests and goals but where reason does
play an important part,
2. A moral psychology and sociology of virtue, which imputes to humans
generally a desire to act benevolently, but where the limits of benevo-
lence requires the establishment of conventional rules of justice, and
3. A moral psychology of traditions and communal conventions, which
bases both social ethics and moral philosophy on historical traditions
and community decisions — these in turn being based on the sentiment
and justice elements of the first two propositions (see Beauchamp,
1991, pp. 256-257).

My approach in this paper acknowledges the seminal influence of Hume,

but is more based on more contemporary philosophical views that resonate
with Hume's vision. The more contemporary proposals are by a subjectiv-
ist (Mackie, 1977), a latter-day emotivist (Gibbard, 1990), and a conserv-
ative sentiment theorist (J.Q. Wilson, 1993). The common part of their
arguments is a simple, and innocuous, sociobiological thesis. (I hasten to
note that I have been critical in some earlier writings of sociobiological
overreachings of the sort found in Richard Alexander's work; see Schaff-
ner, 1978.) The sociobiological argument shows that simple evolutionary
and game-theoretic considerations yield a minimalist account of values
and obligations, but that the minimalist account includes not only "biolog-
ical" but also "social" values, the latter including some community-level
properties, and that further, serve as the basis of a "communitarian" posi-
tion. I know this sounds a bit strange and perhaps a stretch, but indulge me
on this point, temporarily, for the next few paragraphs.
This will be the most controversial part of my argument, but I think it is
sound, and that it can be supported by further evidence of both a direct and
indirect sort. In a brief presentation such as this, I will only be able to
outline the argument. Basically it asks us to consider at least two hypo-
thetical groups of humans (and perhaps even human ancestors, such as the
primates). These are evolving populations in which group membership is
determined by location and interbreedability. In one group, mutation, ge-
netic drift, or possibly some epigenetic variations in behavior/primitive
communication patterns generate cooperative behaviors. In a contrasting
group, no (or fewer) such cooperative behaviors are generated. The as-
370 KENNETH F. SCHAFFNER

sumption is that the group with (more oO the cooperative behaviors will
outproduce the group lacking cooperative behaviors, and thus that cooper-
ation will become evoiutionarily fixed in the descendants.
The cooperative behavior will be grounded in dispositions to behave
cooperatively with other humans, that from a subjective, motivational point
of view will be seen as sentiments, including those of sympathy, empathy
and benevolence. Now this "just-so story" has more going for it than I can
indicate, and has been explored in the group selection debates in evolu-
tionary biology and in the philosophical and game-theoretic literature on
the "Prisoner's Dilemma" that includes in-depth analyses of various strat-
egies examining different advantages of self-interested behavior versus
cooperation. I will not examine these further here, but do want to sketch
what the outcome of this sociobiological process plausibly is. One vision
of that outcome is the kind of sentiment theory we find in James Q. Wilson
(to be distinguished from the sociobiologist Edmund Wilson, though I do
not think he would fundamentally disagree with this vision).
In his book on The Moral Sense, James Q. Wilson argues that "Moral
and political philosophy must begin with a statement about human nature
... — that is a set of traits and predispositions that set limits to what we
may do and suggest guides to what we must do" (1993, p. 235). Wilson
allies himself with Aristotle in this, and asserts that if we are willing to
give up attempts to found a human nature on a single disposition, we can
find a basis on which to construct a moral philosophy that is not only
Aristotelian-like, but which is also consistent with "the findings of mod-
ern science" (p. 236). Wilson's book is an attempt to describe a universal
moral sense that has a number of overlapping components, including sym-
pathy, fairness, and duty (in the sense of honoring obligations). Such a
moral sense, Wilson claims, defending at some length a sociobiological
thesis, arises naturally from the operation of natural selection on groups of
humans and human ancestors.'' This moral sense is complex, and is consti-
tuted by both self-regarding and other-regarding or social virtues. Wilson
writes:

Man is by nature a social animal. Our moral nature grows out of our
social nature. We express sympathy for the plight of others because we
value their company (and so wish to convince them of our companiona-
ble qualities) and because we can feel the pain of others even when not
in their company. We learn to apply the standard of fairness both be-
cause we want to find a cost minimizing strategy for managing joint
activities (such as playing with the same toy) and because we think that
the distribution of things ought to reflect the worth of each beneficiary.
 COMING HOME TO HUME 371

We acquire some measure of self-control both because we value the

good opinion of others and because we value our own self-esteem. We
act on sense of duty both prudently, out of a desire for the praise of
others, and reflexively, in response to an inner command that, though
once spoken by others, now speaks of its own accord and in its own
voice.
The attentive reader will have noticed that in explaining the virtues in
the paragraph above I have combined more or less self-regarding mo-
tives with more or less other-regarding ones (1993, p. 121) [italiciza-
tion of Wilson's "virtues" added].

There are two objections to this sociobiological argument and a sentiment

theory that I want to mention, and suggest why they should be set aside.
The first is that the argument seems to jump the normative-descriptive, or
the "is-ought," divide. This is not the case, I believe, because in terms of a
sentiment theory no such clear division exists. Before you tell me that begs
the question, consider that, further, one can find support for an intertwined
amalgam of the normative — descriptive in thick ethical life as it is expe-
rienced, and from that vantage point see the is-ought distinction as a post-
hoc philosophically reflective construction. This is a point that I take from
Bernard Williams' (1985) analysis of the relation of the normative and
descriptive, an analysis that is complex but that is ultimately strongly
supportive of a human-nature basis for ethical theorizing. I also would
urge you to look at the long argument about this issue that spans much of
Gibbard's 1990 book.
The second objection to my sentiment theory approach is that this
amounts to a kind of emotivism wherein reason does not play a role. This
is false: Sentiments are complex, can conflict, and reason can mediate.
Moreover, sentiments are educable, both in the individual, and historically
across (many) generations. Though the theory is non-objective in the Hume
and Mackie sense, it is not an emotivist account.
Now if you grant me the sociobiological-sentiment theory view, consid-
er what the implications are of this result, that is that individuals who are
disposed to a collection of cooperative behaviors are evoiutionarily more
fit. Those behaviors — seen subjectively as benevolent sentiments — thus
fulfill the functional aspect of Boorse's definition of normal functional
abilities that increase the probability of leaving descendants. But these
sentiments are also moral or evaluative notions, and thus can be the basis
of a prescriptive or valuational element in the now modified and extended
Boorsean account. They can, I would argue, account for why we distin-
guish conditions that produce suffering as diseases. Each of us can "feel
372 KENNETH F. SCHAFFNER

the pain" of others' suffering and react with a disvaiuing response (to the
disease, not the person). There are also some interesting implications of
accepting a sentiment theory like this. Such an account also strongly reso-
nates with a communitarian approach to ethics, as argued by Tom Beau-
champ (1991) in the passage noted above. Partly this is so because the
moral sentiments arose out of — or better, are constitutive of — commu-
nities, and make sense only through inter-individual interactions. Their
implementation and the detailed form they take are the result of long
periods of biological, and, I want to emphasize, social evolution (as well
as individualized experiences from birth on (see Greenspan, 1997)). This
account is, then, not only communitarian, but is also historicist, as Annette
Baier (1987) has pointed out, but this is not necessarily a bad thing. (In
point of fact, I would argue it is a good thing, since it is the best we can
hope for.) It is also not a universalist account, as was Kant's, Mill's, and
maybe (the early) Rawls', but I would also want to urge that with factual
developments conditioning various cultures and the increasing globaliza-
tion of our informational, economic, and political communities, it points
toward a human convergence on a common universalist ethic.
But let us look again at the modified Boorsean concept of health as
seen through this sentiment theory lens. This modification suggests that
health is a moral value, that diseases (but not persons with the diseases)
are disvalued because they involve suffering and functional limitations,
as well as decreased evolutionary fitness, and that restoration to normal
functional ability is morally valuable. Thus this interpretation of a senti-
ment theory triangulates on a view and a list of priorities that Norman
Daniels has argued for using Boorse's notion of normality and an exten-
sion of Rawlsian theory. For his theory, Daniels requires a notion of
health care needs in distinction from preferences, and finds it in a combi-
nation of Boorse's concept of health, that he yokes to Rawls' notion of
equality of opportunity. Health is the absence of disease, and diseases
"are deviations from the natural functional organization of a typical
member of a species" — and here Daniels cites Boorse's works (Daniels,
1985, p. 28). Access to health care given by health-care institutions pro-
vides ways to restore individuals to normal functioning capabilities. And
this, Daniels adds, "can be viewed as a way of keeping the system as close
to possible to the original idealization under which Rawls' theory was
constructed, namely we are concerned with normal, full-functioning per-
sons with a complete lifespan."
This perspective yields a set of rough priorities, now anchored in an
extension of the Boorsean concept of health. Daniels further writes:
 COMING HOME TO HUME 373

Preventive health care institutions can thus be viewed as a first defense

of the idealization: they act to minimize the likelihood of departures
from the normality assumption...[W]e [also] need a second layer of
institutions which corrects for departures from the idealization [of nor-
mality]. These institutions deliver personal medical and rehabilitative
services that restore normal functioning. Similarly not all treatments are
cures, and some institutions and services are needed to maintain persons
in a way that is as close as possible to the idealization. This third layer
of institutions is involved with more extended medical and social sup-
port services for the (moderately) chronically ill and disabled and the
frail elderly. Finally a fourth layer involves health care and related
social services for those who can in no way be brought closer to the
idealization. Terminal care and care for the seriously mentally and phys-
ically disabled fit here, but they raise issues which may not just be
issues of justice. In fact by the time we get to the fourth layer moral
virtues other than justice become prominent (1985, pp. 47-48).

I would argue that the sentiment theory basis of the priorities mentioned
above is, then, actually stronger than one based on Daniels' Rawlsian
extension, since such a sentiment theory provides those moral virtues,
such as benevolence, empathy and charity — Daniels himself refers only
to "beneficence," that Daniels alludes to as needed to account for his
"fourth layer."
In conclusion, the (somewhat) surprising result from this analysis is,
when Boorse is extended to include social sentiments and behaviors, and
this extension is fed back through the Boorsean definition, that a broader
concept of health emerges — broader than is found for example in Boorse's
or in Kass' approaches, but not one as broad as the WHO definition, which
is a more positive concept of health (I will say a bit more about this in the
next paragraph). But this way of arguing also points to a communitarian
and historicist account of value theory (in the Humean sense) that many
will find problematic, if not objectionable, but which I, and I think a small
but growing number of others, do not.
It is important to point out in closing what this restructuring of Boorse
will not do, as well as where it may intercalate with Boorse's most recent
comments on his own theory.^ First, the valuational elements can be quite
complex, as Hume indicated originally, and also as illustrated by Wilson's
multiple sympathies in the quote above (p. 10). But to make the point that
I need to, this value structure need not be complex: it only needs to intro-
duce the concept of sympathizing with a fellow being's suffering as the
basis of a normative dimension to disease. The naturalistic Humean type
374 KENNETH F, SCHAFFNER

of account presented in these pages then jumps the normative-descriptive

divide. True, not every suffering is a disease, and an important aspect of
Boorse's account is that it yields, at least in a prima facie way, a narrower
account than would be permitted by the expansive WHO definition, //we
wish to provide a circumscribed account of disease that includes only
those conditions that match more closely standard nomenclatures of dis-
ease, such as ICDN, we can invoke the biological and psychological di-
mensions of medicine, and stipulate that such suffering must be either
corporeal, or if mental, "unwilled." Boorse's language is that these are
"internal states." My extension of Boorse does, I think, potentially push us
towards a WHO type of definition, but my extension is also historically
conditioned, and thus need not in all societies in all times, embrace home-
lessness, for example, as a disease simpliciter. My account also need not
identify exactly what conditions will count as diseases in any universal —
i.e. timeless or placeless sense, but neither does Boorse's, nor Engelhar-
dt's, nor any general approach. Finally, I want to note that Boorse himself
has recently written that on his purportedly "value-free" account, "one can
build value-laden disease concepts, by adding evaluative criteria, to taste"
(1997, p. 100). What I have tried to show is that valuational elements are
actually constitutive in a natural extension of Boorse's theory, not an add-
on.

NOTES
1. This paper has profited by criticisms from George Khushf, Loretta Kopelman, Lennart
Nordenfelt, and Stan van Hooft. Though t have gone some way toward responding to
their critiques, I do not expect that they will in any full sense embrace the views
developed below.
2. I call utilitarianism "descriptive" on the basis of Mill's argument in his classic text for
the position. I characterize emotivism as "dismissive" based on the distinction be-
tween the Logical Positivists denigration of the cognitive content of ethical expres-
sions and the Humean sentimentalist account favored in this paper.
3. For Boorse's understanding of these and other critiques of his views, and his lengthy
response, see Boorse, 1997.
4. See in particular Engelhardt (1996) on the weaknesses of (biological) evolutionary
theory to provide a complete approach to identifying disease states, and his examples
of sickle cell anemia and post-menopausal osteoporosis (pp. 200-207), but also Boorse's
response in his 1997.
5. Loretta Kopelman suggests this limitation might bias my account by being too restric-
tive. I do not think that this is the case, and believe that one can extend the analysis to
mental disorders without too much change. But such an extension would involve
discussions of the nature of voluntary, involuntary, and nonvoluntary actions, among
other perennial philosophical issues, and cannot be pursued in this paper. Readers
might want to keep Kopelman's point in mind, however.
 COMING HOME TO HUME 375

6. Hume references are to his Treatise in the 1978 Selhy-Bigge edition.

7. Such a moral sense, though it arises naturally from the operation of natural selection
on groups of humans and human ancestors, is not necessarily confined to — in the
sense only directed at — human beings. We naturally extend our considerations of
kindness, etc. to many animals, and I helieve we would similarly do so to as-yet-
unknown species, e.g., were extraterrestrial creatures to he discovered.
8. Several of the points to which I am responding here were raised by George Khushf in
personal correspondence, for which I thank him.

REFERENCES

Baier, A. (1987). 'Hume, the women's moral theorist?' in E.F. Kittay and D.T. Meyers
(eds.). Women and Moral Theory, Rowman & Littlefield Publishers, Totowa; re-
printed in Beauchamp, T. (1991). Philo.sophical Ethics, second edition, McGraw-
Hill, New York, pp. 290-295.
Beauchamp, T. (1991). Philosophical Ethics, second edition, McGraw-Hill, New York.
Boorse, C. (1977). 'Health as a theoretical concept,' Philosophy of Science 44, 542-573.
Boorse, C. (1997). 'A rebuttal on health,' in J. Humber and R. Almeder (eds.) What is
Disease? Humana Press, Totowa, pp. 3-134.
Callahan, D. et al. (1996). 'The goals of medicine: Setting new priorities,' Hastings Center
Report 26 (6), Suppl: S1-S27 + Preface.
Daniels, N. (1985). Just Health Care, Cambridge University Press, Cambridge.
Engelhardt, H.T. (1996). The Foundation of Bioethics, second edition, Oxford University
Press, New York.
Fulford, K.W.M. (1989). Moral Theory and Medical Practice, Cambridge University Press,
Cambridge.
Garrett, D. (1997). Cognition and Commitment in Hume's Philosophy, Oxford University
Press, New York.
Gibbard, A. (1990). Wise Choices, Apt Feelings, Harvard University Press, Cambridge.
Greenspan, S. (1997). The Growth of the Mind, Addison-Wesley, Reading.
Hume, D. (1978). A Treatise of Human Nature, second edition. Clarendon Press, Oxford.
Kass, L. (1985). Towards a More Natural Science: Biology and Human Affairs, Free Press,
New York, esp. pp. 164-174.
Mackie, J. (1977). Ethics: Inventing Right and Wrong, Penguin, London.
Nordenfelt, L. (1987). On the Nature of Health, Reidei, Dordrecht.
Reznek, L. (1987). The Nature of Disease, Routledge and Kegan Paul, London.
Schaffner, K.F. (1978). 'Sociobiology and evolving legal systems: Comments on Richard
A. Alexander's natural selection and societal laws,' in Foundations of Ethics and
its Relation to Science, in D. Callahan and H.T. Engelhardt (eds.), The Hastings
Center, Hastings on Hudson, pp. 291-303.
Schaffner, K.F. and Engelhardt, H.T. Jr. (1998). 'Medicine, philosophy of,' Encyclopedia
of Philosophy, vol. 6, E. Craig (ed.), Routledge, London, pp. 264-269.
WiUiams, B. (1985). Ethics and the Limits of Philosophy, Harvard University Press, Cam-
bridge.
Wilson, J.Q. (1993). The Moral Sense, Free Press, New York.