Content uploaded by Neri Marsili
Author content
All content in this area was uploaded by Neri Marsili on Aug 24, 2019
Content may be subject to copyright.
Forthcoming in Inquiry: an Interdisciplinary Journal of Philosophy!
Immoral lies and partial beliefs
Neri Marsili
Logos – Logic, Language and Cognition Research Group,
University of Barcelona
Abstract: In a recent article, Krauss (2017) raises some
fundamental questions concerning (i) what the desiderata of a
definition of lying are, and (ii) how definitions of lying can account
for partial beliefs. This paper aims to provide an adequate answer
to both questions. Regarding (i), it shows that there can be a
tension between two desiderata for a definition of lying: ‘descriptive
accuracy’ (meeting intuitions about our ordinary concept of lying),
and ‘moral import’ (meeting intuitions about what is wrong with
lying), vindicating the primacy of the former desideratum.
Regarding (ii), it shows that Krauss’ proposed ‘worse-off
requirement’ meets neither of these desiderata, whereas the
‘comparative insincerity condition’ (Marsili 2014) can meet both.
The conclusion is that lies are assertions that the speaker takes to
be more likely to be false than true, and their distinctive
blameworthiness is a function of the extent to which they violate a
sincerity norm.
Keywords: Lying, Credences, Moral Concepts, Definitions, Norms
Introduction: Insincerity and Graded Beliefs
In the philosophical literature on the definition of lying, scholars
agree that the following are necessary conditions for lying:
A. The speaker asserts that p
B. The speaker believes that p is false
Recently, Marsili (2014, 2018) and Krauss (2017) have challenged
this orthodoxy. They both contend that the ‘belief requirement’ (B) #
CONTACT: NERI MARSILI - nerimarsili@ub.edu
is off track, because it is unable to capture graded-belief lies : lies
1
involving a graded belief, i.e. a proposition that the speaker neither
fully believes to be true, nor fully believes to be false. For example,
suppose that Kermit has a credence of 0.25 in (1), so that he
believes that it is probably false that there are chocolate cookies in
the jar:
(1)There are chocolate cookies in the jar
If Kermit tells his friend Elmo that (1), Kermit is lying. But the
‘belief requirement’ (B) prevents the standard definition from
counting (1) as a lie, because Kermit does not fully believe (1) to be
false – he merely believes (1) to be more likely to be false than true.
Marsili and Krauss agree that (B) needs to be revised to capture
graded-belief lies like (1). Marsili’s (2014:162, 2018:176) proposal is
to expand (B) into the ‘comparative insincerity condition’, so as to
capture any credence perceivably lower than 0.5:
The comparative insincerity condition
(CIC) The speaker takes herself to be more confident in the
falsity of p than in its truth
CIC is able to capture graded-belief lies like (1). Nonetheless, Krauss
(2017) has recently alleged that CIC is mistaken, because it fails to
‘account for the damage lying does’. His alternative proposal is that
an assertion is a lie only if it is expected to make the addressee
‘epistemically worse-off’; that is, only if it satisfies the ‘worse-off
requirement’:
The worse-off requirement
(W-O) The expected epistemic damage to the audience, with
respect to p, by the speaker’s lights, conditional on the
audience trusting her with respect to p, at all, is greater
than 0
For Marsili (2016), however, there is a further reason to reject (B), namely that it
1
fails to capture some insincere promises: those that you believe you will likely keep,
despite your intention to violate them.
$ of $2 13
In simpler words, in asserting that p you lie only if you expect your
audience’s credence in p to become more inaccurate than the one
they previously held (if they trust your assertion that p). Applied to
our example, this means that (1) is a lie whenever Kermit asserts (1)
and expects Elmo to update his credence in p to a value that is more
distant from 0.25 (which is Kermit’s standard of accuracy) than
Elmo’s current credence in p. For instance, Kermit is lying if he
expects Elmo to revise his credence in (1) from 0.4 to 1, because 1 is
further away from 0.25 than 0.4.
Krauss prefers W-O to CIC for one crucial reason: he takes the
former, but not the latter, to be able to account for the ‘distinctive
blameworthiness’ of lying. He shows this by means of a
counterexample:
Imagine a case in which the speaker is just slightly more confident that p
is true than false – say she has 0.51 credence in p. And, further, imagine
that the speaker knows that her audience is agnostic – that the audience
has 0.5 credence in the proposition. […] Imagine that the audience moves
from 0.5 credence to 0.8 credence. If the speaker’s credence is 0.51, then
the speaker will expect the audience to suffer epistemic damage. If the
blameworthiness for lying is grounded in expected epistemic damage,
then this speaker is blameworthy in exactly the same way as liars are,
even if, according to both the orthodox position and Marsili’s proposal,
she hasn’t lied.
The effectiveness of this counterexample is conditional on two
assumptions: (i) that the primary desideratum of a definition of lying
is to track the distinctive blameworthiness of lying, and (ii) that this
blameworthiness is best characterised in terms of expected epistemic
damage, as defined by the worse-off requirement. These assumptions
are not uncontroversial, as they take a firm stance on two
fundamental issues in the philosophy of lying: what the primary
desiderata of a definition of lying are (for assumption (i)), and what
makes lying blameworthy (for assumption (ii)). If these assumptions
are correct, the ones underlying the current philosophical debate on
lying must be wrong, as both (i) and (ii) are overwhelmingly held to
be false. In this paper, I vindicate a ‘traditional’ stance on these
issues, and show that, at closer inspection, both assumptions are
indeed wrong. Against assumption (i), I argue that capturing the
$ of $3 13
distinctive blameworthiness of lying cannot be the only desideratum
of a definition of lying. Against assumption (ii), I show that the
worse-off requirement (W-O) fails to track the distinctive
blameworthiness of lying, which is better characterised in terms of
the violation of a sincerity norm.
1. The Desiderata of a Definition of Lying
I will start by dealing with the first assumption: that the primary
desideratum of a definition of lying is to track the distinctive
blameworthiness of lying. To avoid misunderstandings, let me stress
that this assumption should not be confused with the view that all
lies are, all things considered, morally wrong. Krauss is merely
assuming that a good definition should track the fact that lies are
blameworthy in the same distinctive way (e.g. in virtue of the
epistemic risk that they impose on the audience). This still allows for
single acts of lying to be ‘all things considered’ permissible, whenever
sufficient countervailing considerations apply (Margolis 1963).
The idea that a definition should track the distinctive
blameworthiness of lying is not entirely new. Some authors before
Krauss have assumed that a negative moral evaluation is part of the
meaning of the word ‘lie’ (Margolis 1963, cf. also Williams 1985,
140), so that in defining lying one should also explain what makes
lying blameworthy (Grotius 1625, book III.I.XI.1) . This is
2
nonetheless a minority view, and in contemporary scholarship the
prevailing assumption is that the key desideratum of a definition of
lying is its ‘descriptive accuracy’: a good definition should match our
(morally neutral) intuitions about particular cases, capturing all and
only those utterances that we call lies. The task of defining what a
lie is and explaining why (ceteris paribus) lying is wrong are on this
conception two separate tasks (Kemp and Sullivan 1993, Fallis 2009,
Carson 2010:13, Mahon 2015). In assuming that a definition should
first of all capture the distinctive blameworthiness of lying, Krauss is
In his De Jure Belli ac Pacis, Grotius (1625, III.I.XI.1) declares that his intent is
2
not to define lying in general, but the notion of a ‘Lye strictly taken, as it is naturally
unlawful’ (that is, wrong or blameworthy). On his view, this ‘strict’ notion of lying
necessarily involves the ‘Violation of a real Right’ (that is, an undefeated right to be
told the truth).
$ of $4 13
thus departing from the methodological conventions in the literature,
setting a different priority for his proposal: to explain what makes
lying prima facie wrong.
Let us assume that Krauss has good motivations for introducing this
new desideratum of, as we may call it, ‘moral import’. The crucial
problem with its introduction is that the two resulting desiderata
(moral import and descriptive accuracy) can in principle come apart
(i.e. impose inconsistent constraints on the definition); when they do,
a morally interesting definition will not be descriptively accurate.
The worse-off criterion (W-O) proposed by Krauss can be helpful to
illustrate how moral import and descriptive accuracy can come apart.
Suppose that we share Krauss’ intuition that the distinctive moral
wrong in lying is that the liar imposes an epistemic threat on the
dupe, and that W-O is successful in tracking this distinctive kind of
blameworthiness. This would not yet guarantee that a definition
incorporating W-O is descriptively accurate. To see this, consider the
following counterexample (inspired by Benton 2018).
Suppose that (i) Kermit tells Elmo that (1) is false (he tells Elmo
that there are no chocolate cookies in the jar), (ii) Kermit is
maximally certain that (1) is true (Kermit is certain that there are in
fact some chocolate cookies in the jar), and (iii) Kermit is aware that
Elmo is already maximally certain that (1) is true (he knows that
Elmo is already convinced that what he just said is true). While
Kermit is clearly lying , he is not attempting to modify Elmo’s degree
3
of belief in (1) – he is merely providing a further (testimonial) reason
in support of that belief. Since the expected epistemic damage (from
the speaker’s perspective, and if the hearer trusts him) is not greater
than 0, W-O fails to classify this as a lie. More generally, this sort of
counterexamples (cf. Benton 2018 for further ones) shows that W-O
is not descriptively accurate – it is too narrow, because it fails to
provide a criterion that captures all lies.
Virtually every definition on the market converges on this prediction, including
3
those that require an intention to deceive. Note that although Elmo already believes
that p is false, Kermit is attempting to deceive Elmo on most standard accounts of
deception, as he brings about new (testimonial) evidence for that belief, contributing
causally to Elmo’s continuing to have that inaccurate belief (cf. Chisholm and
Feehan 1977, 144, Fuller 1976; Mahon 2007, 186-7, 189–90).
$ of $5 13
This seems to be a problem for Krauss’s view, for he holds true (a)
and (b), which seem to be incompatible with (c), that we have just
proved to be true:
(a) Lies are blameworthy in the same way
(b) Lies are blameworthy only if [they meet the worse-off
requirement]
(c) There are some lies that [do not meet the worse-off requirement]
Since (a), (b) and (c) cannot all be true, one of these claims needs
to be abandoned. If the desideratum of moral import needs to play
some role in the definition of lying, we cannot abandon (a). The
conclusion is that (b) must be abandoned.
This points out to a more general problem: that descriptive accuracy
cannot be sacrificed at the expenses of moral import. The schema (a-
b-c) generalises to any account of lying, once the content of the
square brackets is replaced by an alternative requirement for the
definition of lying. Note that claim (c) here represents the negation
of the desideratum of descriptive accuracy, whereas condition (a)
represents the desideratum of moral import. The schema shows that
if descriptive accuracy is not met (that is, if (c) is true), moral
import cannot be achieved (that is, (b) must be abandoned)– as long
as we accept (a). Generalising, the schema shows that correctly
characterising the moral import of lying is conditional on a
definition’s descriptive accuracy – whenever a definition fails to
4
Note, however, that the ‘only if’ in (b) only captures the necessity leg of the
4
desideratum of moral import, and (c) only represents a challenge to the necessity leg
of the definition. This because my main focus is W-O, and Krauss only claims that
W-O is a necessary condition for lying (and for being blameworthy in the distinctive
way that liars are). To generalize the schema so that it applies to full definitions, we
should replace the ‘only if’ in (b) with a biconditional, and weaken (c) so as to allow
for counterexamples to sufficiency.
$ of $6 13
capture all lies, it also fails to capture the feature that makes lies
blameworthy in the same distinctive way .
5
We are now in a position to determine what is wrong with
assumptions (i) and (ii). The problem with (i), the view that the
primary desideratum of a definition is tracking the ‘moral import’ of
lying, is that meeting this desideratum is conditional on meeting
‘descriptive accuracy’. As illustrated by the (a)-(b)-(c) scheme, any
definition of lying that fails to meet descriptive accuracy also
inevitably fails to capture the distinctive blameworthiness of this
linguistic phenomenon. Against assumption (ii), W-O neither offers a
good necessary condition for defining lies (since it is subject to
counterexamples), nor it is able (for the same reason) to track their
distinctive blameworthiness.
To be sure, to deny these two assumptions is not to say that it is
impossible to identify a criterion that is able to meet both desiderata.
Clearly, there could be an alternative way to capture a morally salient
feature that all lies have (thereby meeting ‘moral import’), and it
may turn out that this feature is shared by all lies (thereby meeting
‘descriptive accuracy’). But in order to determine whether such a
criterion exists, we need to test it against both desiderata (pace to
Krauss 2017:2,4,5). In what follows I argue that, at closer inspection,
One way to resist this conclusion is to argue that definitions should only aim to
5
capture a morally uniform phenomenon, that can differ significantly to our ordinary
concept of lying. The aim of this radical ‘revisionist’ project would be to identify a
concept that has moral significance, lying*, that may or may not coincide with our
ordinary concept of lying. If it does coincide, the definition also provides an analysis
of our ordinary concept. But if it does not, the definition of lying* thus obtained
would still be illuminating: for instance, it could be useful for moral theorising, or for
law-making purposes. This alternative project is certainly coherent, and possibly
worth pursuing. For my purposes, it is sufficient to point out that this project is so
radically different from the one currently pursued by scholars working on the
definition of lying that it cannot be regarded as a continuation of it. Authors
engaging in the current debate explicitly aim to characterise the ordinary concept of
lying (as it is employed by laypeople in their reasoning and talking about lying), as
opposed to an artificial concept of lying* – within a common effort to engage in
conceptual analysis, rather than conceptual engineering (cf. Fallis 2009, Mahon
2015). Lack of concern for ordinary use is thus incompatible with intervening within
this pre-existing debate (which is, quite uncontroversially, what Krauss 2017 aims to
be doing).
$ of $7 13
the ‘comparative insincerity condition’ CIC can help to meet this
challenge: it identifies a concept that coincides with our ordinary
conception of lying, and tracks the distinctive blameworthiness of this
phenomenon.
2. Accuracy with Blameworthiness
2.1. Rescuing the Comparative Insincerity Condition
I will begin by reconsidering why CIC was rejected in the first place.
Krauss’ criticism of CIC relied on the assumptions that we just
rejected: that a definition of lying only needs to meet the
desideratum of moral import, and that having W-O as a necessary
condition is the only way to meet this desideratum. Now that we
have established both these premises to be misguided, the
counterexample on the ground of which Krauss rejects CIC loses its
intuitive pull.
In the counterexample to CIC discussed in §1, a speaker asserts that
p while having a credence of 0.51 in p. Unless we assume that a
definition only has to track the distinctive blameworthiness of lying,
and adopt the worse-off requirement as the criterion to identify this
blameworthiness (but we have seen compelling arguments not to),
there is no strong reason to think that this assertion is either a lie or
not a lie. Intuitions about this sort of cases are not straightforward,
and do not warrant a preference for either CIC or W-O. But suppose
that we are forced to give a polar verdict, and to establish that either
this is a lie or it is not. Intuitions may vary, but the fact that the
speaker is more confident in the truth of p than in its falsity at least
hints at the fact that this is not a lie. CIC is able to make sense of
both these observations. W-O, by contrast, has no resources to
acknowledge that the example is a borderline case; nor is it able to
acknowledge, like CIC, that the more we move from these cases of
uncertainty (around 0.5) to certainty (0 or 1), the sharper will be our
intuitions about whether a given utterance is a lie (cf. Marsili
2018:176).
At any rate, if the descriptive accuracy of a given account is best
measured against straightforward cases, this example bears limited
argumentative weight. Crucially, when we consider straightforward
$ of $8 13
cases of lying, CIC makes the right verdict in every scenario in which
other definitions fail. Unlike W-O, it avoids Benton’s
counterexamples (as discussed in Benton 2018:3, fn5); and unlike the
standard ‘belief requirement’ (B), it is able to rule in graded-belief
lies. In sum, CIC fares better than its rivals in terms of descriptive
accuracy.
2.2. Blameworthiness as Norm-Violation
Is CIC also able to track the distinctive blameworthiness of lying? I
think that this question can be answered positively. It can be argued
that such blameworthiness resides in a feature identified by CIC: the
(more or less severe) violation of a norm prescribing speakers to be
sincere. Like the standard insincerity belief requirement (B), CIC
captures a set of acts that have a morally salient feature in common
– they all violate a sincerity norm. But improving on (B), CIC
acknowledges the intuitive fact that a lie can be ‘a more or less
severe violation’ of such a norm (Marsili 2014, 2018): one thing is to
present as true a proposition that you are certain to be false, another
to present as true a proposition that you merely believe more likely to
be false than true. In other words, CIC represents an improvement on
the tradition both in terms of ‘descriptive accuracy’ and in terms of
‘moral import’.
Reinterpreted as characterising the wrongness of lying in terms of
norm-violation, CIC places itself in a ‘classic’ tradition in the
literature on the morality of lying. What is wrong with lying is here
understood in deontological terms – there is a rule that lying
infringes, and its blameworthiness is a function of such infringement.
By contrast, Krauss’ account identifies the blameworthiness of lying
in terms of its expected harmful consequences, falling rather under a
consequentialist tradition.
This might be seen as a sign that there is something counterintuitive
about CIC. Prominent proponents of deontological accounts like
Augustine and Kant also subscribed to ‘absolutism’ about lying: the
view, often judged to be counterintuitive, that the impermissibility of
lying is exceptionless (cf. Augustine [DM]; [CM], Kant [G]; [LE];
[RL]). But accepting a deontological story about the wrongness of
lying does not commit one to absolutism. One can still maintain that
$ of $9 13
while all lying is wrong qua violation of a sincerity norm, any given
act of lying can be nonetheless overall morally permissible; to put it
in slightly different terms, a plausible version of this view is that lying
is prima facie wrong (cf. Ross 1930). On this view, the fact that you
say something you believe more likely to be false than true is a
defeasible reason to classify your action as morally reprehensible.
Furthermore, saying that lying is wrong in virtue of the violation of a
sincerity norm does not entail that the wrongness of lying is
exhausted by such violation. This view allows for other criteria of
moral evaluation to be salient when we make moral judgments about
lying (cf. Stokke 2017). For instance, the expected deceptiveness of a
lie will typically be a salient dimension of evaluation, as hardly
anyone denies that lying typically involves intended deception.
3.3 A Reconciliation: Lies and Deception
We can conclude by devising a partial reconciliation between Marsili’s
and Krauss’ view. When it comes to a descriptively accurate
definition of lying, lying is best understood as the act of asserting
something insincerely, where insincerity is captured by CIC. All lies
are prima facie wrong qua lies, to the extent (captured by CIC) that
they all violate a norm of sincerity– the greater the violation, the
more severe the blameworthiness that arises from such violation.
Furthermore, most lies are prima facie wrong qua attempts to
deceive; arguably, this kind of blameworthiness is a function of the
extent to which they meet W-O: the greater the epistemic risk they
impose on the hearer, the greater the blameworthiness of the attempt
to deceive . This gives us a two-fold account of the blameworthiness
6
of lying: in terms of the credences they necessarily misrepresent, and
of the inaccurate credences they typically aim to induce.
While there is growing consensus in the literature that lying does not necessarily
6
involve an intent to deceive, some philosophers still subscribe to this view (e.g.
Faulkner 2007; Lackey 2013). If they are right, then all lies are also blameworthy qua
attempts to deceive. In either case, in light of Benton’s (2018) counterexamples, W-
O still needs to be refined to successfully track the blameworthiness of each and
every act of attempted deception.
$ of $10 13
Conclusions
This article has proposed a solution to three fundamental questions
in the philosophy of lying: (i) establishing the primary desiderata of a
definition lying, (ii) defining what lying is, and (iii) characterising its
distinctive blameworthiness.
In relation to the first issue, I argued (against Krauss and a tradition
that traces back to Grotius) that offering a definition of lying and
characterising its moral import are two independent tasks, and that
achieving the latter is conditional on achieving the former. When it
comes to evaluating whether a definition of lying is correct, we
should thus look at a definition’s descriptive accuracy rather than its
ability to explain the distinctive way in which lies are blameworthy,
since only by meeting the former desideratum we can meet the latter.
These considerations have important consequences for the ongoing
philosophical debate about what lying is, and what makes it
blameworthy. Once Krauss’ mistaken assumptions about the
desiderata for the definitions of lying are abandoned, it becomes clear
that his critique of the ‘comparative insincerity condition’ is
misguided. This in turn means that we can after all offer a fine-
grained definition of lying that accounts for partial beliefs – for which
no options are otherwise left available, given that neither Krauss’
‘worse-off requirement’ nor the traditional ‘belief requirement’ are
able to draw the right distinctions about graded-belief lies.
Finally, I have argued that CIC has the resources to track the
distinctive blameworthiness of lying, it in terms of a (more or less
severe) violation of a sincerity norm. Endorsing this explanation of
the moral import of lying is compatible with acknowledging that also
W-O captures a morally salient dimension of blameworthiness: the
(higher or lower) epistemic damage that the liar typically aims to
impose on the dupe. In advancing this partial reconciliation between
two competing views about graded-belief lies, this paper has offered
a novel, bipartite characterisation of lying and its distinctive
blameworthiness: in terms of the norms that it necessarily violates,
and in terms of the epistemic risks that it typically imposes on the
hearer.
$ of $11 13
References
Augustine. [DM], De Mendacio, ‘On Lying,’ M. S. Muldowney
(trans.) (51–110)
Augustine. [CM], Contra Mendacio, ‘Against Lying,’ H. B. Jaffee
(trans.) (121–179), in R. J. Deferrari (ed.)!Fathers of the
Church!(Volume 16:!Treatises on Various Subjects), New York:
Fathers of the Church, 1952
Carson, Thomas L. 2010. Lying and Deception. Oxford University
Press.
Chisholm, Roderick M., and Thomas D Feehan. 1977. The Intent to
Deceive. Journal of Philosophy 74 (3): 143–59.
Fallis, Don. 2009. What Is Lying? Journal of Philosophy 106 (1): 29–
56. internal-pdf://0.0.4.12/FALWIL.html.
Faulkner, Paul. 2007. What Is Wrong with Lying? Philosophy and
Phenomenological Research LXXV (3): 535–57.
Fuller, Gary. 1976. Other-Deception. The Southwestern Journal of
Philosophy 7 (1): 21–31.
Grotius, Hugo. 1625. The Rights of War and Peace. Edited by Jean
Barbeyrac (2005 edition). Indianapolis: Liberty Fund. https://
oll.libertyfund.org/titles/grotius-the-rights-of-war-and-peace-2005-
ed-vol-3-book-iii.
Kant, I. [G].!Groundwork of the Metaphysics of Morals, in!Immanuel
Kant, Practical Philosophy, edited by A. W. Wood and M. J.
Gregor, Cambridge: Cambridge University Press, 1996.
Kant, I. [RL]!On a supposed right to lie from philanthropy,
in!Immanuel Kant, Practical Philosophy, Cambridge: Cambridge
University Press, 1996.
Kant, I. [LE]!Lectures on Ethics, P. Heath (trans.), P. Heath and J.
B. Schneewind (eds.), Cambridge: Cambridge University Press,
1997.
Kemp, Kenneth W., and Thomas Sullivan. 1993. Speaking Falsely
and Telling Lies. Edited by Michael Baur. Proceedings of the
American Catholic Philosophical Association 67: 151–70. doi:
10.5840/acpaproc19936712.
Krauss, Sam Fox. 2017. Lying, Risk and Accuracy. Analysis 73: 651–
59. doi:10.1093/analys/anx105.
Lackey, Jennifer. 2013. Lies and Deception: An Unhappy Divorce.
$ of $12 13
Analysis 73 (2): 236–48. doi:10.1093/analys/ant006.
Mahon, James Edwin. 2007. A Definition of Deceiving. International
Journal of Applied Philosophy 21 (2).
Mahon, James Edwin. 2015. The Definition of Lying and Deception.
Stanford Enciclopedia of Philosophy.
Margolis, Joseph. 1963. “Lying Is Wrong” and “Lying Is Not Always
Wrong.” Philosophy and Phenomenological Research 23 (3): 414–
18.
Marsili, Neri. 2016. Lying by Promising. International Review of
Pragmatics 8 (2). Brill: 271–313. doi:
10.1163/18773109-00802005.
Marsili, Neri. 2014. Lying as a Scalar Phenomenon. In Certainty-
Uncertainty – and the Attitudinal Space in Between, edited by S.
Cantarini, W. Abraham, and E. Leiss, 153–73. Amsterdam: John
Benjamins Publishing Company. doi:10.1075/slcs.165.09mar.
Marsili, Neri. 2018. Lying and Certainty. In The Oxford Handbook of
Lying, edited by Jörg Meibauer, 169–82. Oxford: Oxford
University Press. doi:10.1093/oxfordhb/9780198736578.013.12.
Ross, W. D. 1930. T h e R i gh t a n d t he G o o d . d o i:
10.1093/0199252653.003.0001.
Stokke, Andreas. 2017. Lies, Harm, And Practical Interests.
Philosophy and Phenomenological Research, no. 2007: 1–17. doi:
10.1111/phpr.12439.
Williams, Bernard Arthur Owen. 1985. Ethics and the Limits of
Philosophy. Routledge.
$ of $13 13
