Received: 17 October 2018 | Revised: 19 February 2019 | Accepted: 14 April 2019

DOI: 10.1111/rati.12236

ORIGINAL ARTICLE

Hope as an irreducible concept

Claudia Blöser

Department of Philosophy, Goethe

University Frankfurt, Norbert‐Wollheim‐
Platz 1, 60629, Frankfurt am Main, Germany
Correspondence
Claudia Blöser, Department of Philosophy,
Goethe University Frankfurt, Norbert‐
Wollheim‐Platz 1, 60629 Frankfurt am
Main, Germany.
Email: bloeser@em.uni-frankfurt.de
Abstract
I argue for a novel answer to the question “What is hope?”.
On my view, rather than aiming for a compound account, i.e.
analysing hope in terms of desire and belief, we should under-
stand hope as an irreducible concept. After criticizing influen-
tial compound accounts of hope, I discuss Segal and Textor's
alternative of describing hope as a primitive mental state.
While Segal and Textor argue that available developments of
the standard definition do not offer sufficient conditions for
hope, I question the deep‐seated idea that desire and belief
are even necessary conditions for hope. My suggestion is that
we should take seriously the fact that we hope in a great vari-
ety of ways and should question the search for elements that
are common to all cases. A promising alternative follows the
Wittgensteinian idea that cases of hope are related in terms of
family resemblance, i.e. are multiply realizable on the ontologi-
cal level while falling under the non‐definable concept of hope.

KEYWORDS
belief, desire, family resemblance, hope, moral psychology

In this article, I present a new answer to the question “What is hope?”. On my account, we should understand hope
as a concept that is not reducible to desire and belief. The first section presents the most common approach,
which takes as its starting point the standard definition, according to which hope is analysed in terms of a desire for
an outcome and a belief that the outcome is possible. This definition has been convincingly criticized as insuffi-
cient, and the prevalent consequence of this criticism is the search for further conditions that complement the
account. I call those accounts that analyse hope in terms of other concepts or states “compound accounts”. By way
of an exception, Segal and Textor (2015) take a radically different route and describe hope as a primitive mental
state that is not reducible to other mental states (section 2).1 On my view, Segal and Textor's account points in the

1
The only other person who discussed aspects of a simple state account even earlier than Segal and Textor is Darrel Moellendorf (2006), who ulti-
mately rejects it.

Ratio. 2019;00:1–10. wileyonlinelibrary.com/journal/rati © 2019 John Wiley & Sons Ltd | 1

2 | BLÖSER

right direction. However, I discuss problems with their approach that motivate my own position. Segal and Textor
argue that available developments of the standard definition do not offer sufficient conditions for hope. I offer a
more fundamental critique of the standard definition, questioning the deep‐seated idea that desire and belief are
even necessary conditions for hope (section 3). In the fourth section, I argue against Segal and Textor's idea that
hope can be characterized by its unique motivational role, which speaks against understanding it as a primitive
mental state. Section five introduces my own suggestion: We should pursue the Wittgensteinian idea that cases
of hope are related in terms of family resemblance, i.e. they are multiply realizable on the ontological level, while
all falling under the irreducible concept of hope.

1 | TH E O RTH O D OX D E FI N ITI O N , IT S C R ITI C S A N D AU G M E NTE D

CO M P O U N D ACCO U NT S

Despite the variety among the available accounts of the nature of hope, most are based on what has been called
the “standard account” or “orthodox definition”, according to which “hope that X” is constituted by a desire for X
and the belief that X is possible. 2 This standard definition has been convincingly criticized as insufficient. Luc
Bovens was the first to make this point, arguing that “[t]here is no conceptual confusion in affirming one's desire
for some state of the world and one's belief that this state may or may not come about, while denying that one is
hoping that it would come about” (Bovens, 1999, p. 674). Many philosophers agree with this claim and illustrate it
with examples, comparing two people (or one person at two points in time) who equally desire an outcome and
who both believe in its possibility (but not its certainty) while nonetheless differing in crucial respects, e.g. in
terms of behaviour, affective outlook, and self‐interpretation. Ariel Meirav picks up what has become the most
famous of these examples, which is based on Frank Darabont's film The Shawshank Redemption (Meirav, 2009,
p. 222).3 Two prisoners, Andy and Red, both desire to be free, and both believe that there is a (very small) chance
that they will escape the prison. Yet Andy hopes for freedom, while Red despairs over his situation. Thus, the
standard definition does not have the resources to distinguish between those who hope and those who despair.
Adrienne Martin's cancer research example has a similar structure: Two terminally ill cancer patients, Alan and
Bess, have equally strong desires for a “miracle cure”, and both know that there is a less than 1 per cent chance
that an experimental drug will be successful. They both enrol in the drug trial, but whereas Bess really “hopes to
be the 1 percent”, Alan focuses on how slim his chances are, and thus his hope is weaker (Martin, 2013, p. 15).
Similarly, Darrel Moellendorf imagines a political activist who, after years of engaging in struggles for justice, gives
up her activism. She still desires a just society as much as before and continues to believe that it is possible; how-
ever, she is no longer stirred to action. She is demoralized (Moellendorf, 2006, p. 419). Segal and Textor share
Moellendorf's intuition that hope affects motivation in ways that cannot be explained by the mere presence of
desire and belief. They invoke the example of two mountaineers, Reinhold and Hillary, who are climbing a moun-
tain in dire weather conditions. By stipulation, they share the relevant desires and beliefs, but their levels of hope
differ, which has implications for what they are likely to do: “Reinhold hopes that he will make it to the summit,
hence he is likely to go; Hillary has lost heart, hence, he is likely to return. Hope can break the deadlock” (Segal &
Textor, 2015, pp. 209f.).
These counterexamples show that the standard definition does not capture what hope is. Note, however,
that different critics are concerned with different features that they take to be definitive of hope: Meirav focuses
on hope's “characteristic positive outlook” (Meirav, 2009, p. 218); Martin emphasizes that a hoping person sees
herself as justified in pursuing certain activities; Moellendorf, Segal and Textor are concerned with motivation. I
will return to this point in section 4. Here, I merely want to agree with the general line of criticism of the standard

2
For the labels, see Meirav (2009, p. 217) and Martin (2013, p. 11). Downie (1963) is an example of a proponent of the standard definition.
3
The example was originally introduced into the debate by Luc Bovens to make a different point (Bovens, 1999, pp. 668f.).
BLÖSER | 3

definition: it does not offer sufficient conditions for hope. This result either leads to the claim that the standard
definition must be augmented or motivates the proposal that hope is entirely different from desire and belief, and
hence irreducible to them.
Most accounts pursue the first possibility.4 Segal and Textor criticize three prominent attempts to do so (put
forward by Luc Bovens, Philipp Pettit and Adrienne Martin) in order to motivate their alternative view that hope
is an irreducible mental state. I will follow this structure and briefly voice criticisms of existing compound ac-
counts, before criticizing Segal and Textor's account as well.
Luc Bovens suggests that “[h]oping is just having the proper belief and desire in conjunction with being en-
gaged to some degree in mental imaging” (Bovens, 1999, p. 674). He illustrates this with the example of Sophie,
who arrives late to a party and asks the host self‐confidently whether he had hoped that she would come. Bovens
claims that hoping for her to come would require devoting “at least some mental energy” (Bovens, 1999, p. 674),
such as checking whether Sophie is amongst the newly arriving guests. However, even though Bovens is right that
some mental imaging is typically present when one hopes, the standard definition has resources to account for
this. Desire can entail mental imaging – if not every desire, then desires of a certain intensity.5 More importantly,
mental imaging does not allow us to distinguish between cases of hope and cases of despair. It is plausible to as-
sume that both Andy and Red devote mental energy to the prospect of being free, while one hopes and the other
despairs.6
Philipp Pettit criticizes the standard definition as incomplete because it cannot explain why hope is able to
sustain our actions in situations of great uncertainty. Such “substantial” hope, according to Pettit, involves cogni-
tive resolve. A person T who hopes that p “has resolved to act as if p will be the case (or is highly probable) because
T wants to act in the face of his low degree of conviction” (Pettit, 2004, p. 159). Various objections have been
raised against this proposal. For example, Segal and Textor claim that because one cannot choose to hope but one
can choose to resolve to act, hope cannot be a resolve to act (Segal & Textor, 2015, p. 218). While one might want
an argument for the claim that one cannot (ever) choose to hope, I agree with Segal and Textor in general: We
cannot choose to hope in the same way (i.e. directly, always) that we can choose to act (even though it may be
possible to cultivate hope), and thus a resolve to act cannot be equated with hope.7
On Adrienne Martin's account, “a hopeful person stands ready to offer a certain kind of justificatory rationale
for her hopeful plans, thoughts, and feelings” (Martin, 2013, p. 36). More specifically, a hopeful person sees ‘the
probability she assigns the hoped‐for outcome […] as licensing her to treat her desire for the outcome and the
outcome's desirable features as reasons to engage in […] forms of planning, thought, and feeling’, and she actually
treats “her desire and the outcome's desirable features as reasons to engage in said forms of planning thought, and
feeling” (Martin, 2013, p. 35). Her position has been criticized as being excessively reflective (Segal & Textor, 2015,
p. 220f.). In particular, she cannot account for “recalcitrant hopes”,8 i.e. hopes that we do not see as justified and
that we would prefer to give up. Martin holds that these are “not full‐fledged hope, but potentially worth calling
‘hope’ anyway” (Martin, 2013, p. 62). However, if a person says “I still hope that my ex‐husband will return, even
though I think I should rather give up this hope and stop thinking about him”, we do not hesitate to attribute gen-
uine hope to that person. Martin's conception of hope is too demanding in viewing the rational activity of taking
one's desire (or attraction) as a reason as partially constitutive of genuine hope.9

4
See e.g. Bovens (1999), Martin (2013), Meirav (2009), Milona (2018), and Kwong (2019).
5
See e.g. Meirav (2009, p. 221), Martin (2013, pp. 17f.), and Segal & Textor (2015, p. 212).
6
For the same point, see Kwong, 2019, fn2.
7
See Han‐Pile (2017) for the same criticism.
8
For the same criticism, see Milona & Stockdale (2018, pp. 208f.) and Han‐Pile (2017).
9
The same objection applies to Moellendorf's account, which is similar to Martin's. He holds that, in combination with desire and belief, “[t]he taking
of one's beliefs and desires as reasons for acting is a necessary condition for hoping” (Moellendorf, 2006, p. 420).
4 | BLÖSER

Note that Martin does not want to offer an analysis in terms of necessary and sufficient conditions. Instead,
her “syndrome analysis” amounts to a “set of paradigmatic marks of the analysandum” (Martin, 2013, p. 6). While
the complete set of features characterizes “hope in the fullest sense”, “many of the elements are not necessary
for hope” (Martin, 2013, p. 62). However, I think the criticism is apt nonetheless. Martin emphasizes that her set
of conditions differs from a “random collection of things” (Martin, 2013, pp. 6f.) because it contains a unifying
element: incorporating one's attraction to the outcome into one's agency and hence seeing oneself as justified in
pursuing hopeful activities. It is precisely this unifying element that leads to an excessively reflective conception
of hope. That is, whereas definitions can be criticized for excluding certain cases from being hope at all, Martin's
syndrome account can be criticized for excluding certain cases from being genuine hopes, or hopes in the fullest
sense. Her unifying element construes genuine hope as excessively reflective.
There is an ever‐increasing number of compound accounts that take the standard definition as their starting
point.10 My own account does not depend on rejecting every new account (see section 3). The brief survey of
criticisms of compound accounts should simply raise doubts about the possibility of analysing hope in terms of
simpler components. Let us consider in the following section an alternative to compound accounts: Segal and
Textor's proposal that hope is irreducible in the sense of being a primitive mental state.

2 | S EG A L A N D TE X TO R ' S ACCO U NT O F H O PE A S A PR I M ITI V E

M E NTA L S TATE

Segal and Textor claim that “hope can neither be defined in terms of belief and desire nor any other mental state:
it is a primitive mental state” (Segal & Textor, 2015, p. 207). Their argument can be summarized as follows:

1. If hope has a distinctive feature that distinguishes it from desire, and this feature is not likely to be
accounted for by compound accounts, this speaks in favour of the irreducibility of hope.
2. Hope has distinctive motivational force independent of desire.
3. This feature has not been successfully captured in existing compound accounts of hope.
4. Most philosophers have given up on the project of providing definitions of mental states in general. Instead,
mental states are characterized by their functional role in relation to other mental states (Segal & Textor, 2015,
p. 207).
5. Therefore, we should understand hope as irreducible and describe its functional role.

Premise (4) is a reason to renounce the project of finding a reductive account of hope only for those who are inclined
towards functionalism regarding mental states. However, those who are engaged in the project of finding a compound
account of hope must rely on premises (2) and (3) in order to be convinced of the irreducibility of hope.
Premise (1) is an implicit premise for Segal and Textor and is supposed to clarify the argumentative role of
premises (2) and (3). As already mentioned, Segal and Textor establish premise (2) by way of the example of the
two mountaineers: What explains the difference in motivation between Reinhold and Hillary is that Reinhold
hopes and Hillary does not. Segal and Textor want to account for hope's motivational force not in terms of a third
condition of a compound account but by providing a functionalist description. On their account, describing hope's
relation to other mental states like desires and beliefs should capture hope's motivational force. Their specific
proposal is the following:

This aspect of hope [its motivational force, CB] can be captured in one general characterization:
suppose G is a possible desirable outcome of an attempted action A. Then, keeping fixed the

10
Recent accounts include Calhoun (2018), Milona (2018), and Kwong (2019).
BLÖSER | 5

subjective value of G and subjective likelihood of G, given A, the extent of a subject’s motivation
for doing A depends on the extent to which he hopes that if A, then G (Segal & Textor, 2015, p. 222).

There are two problems with Segal and Textor's account. First, their argument for the irreducibility of hope hinges
on the idea that the distinctive feature of hope is not likely to be accounted for by compound accounts. Premise (3)
carries a lot of argumentative weight: In order to be convincing in the long run, Segal and Textor would have to ex-
amine every new candidate for a compound account and reject it. In order to avoid this problem, I argue in the next
section against the idea that belief and desire are necessary components of hope, which amounts to a refutation of all
compound accounts. The second problem concerns Segal and Textor's specific proposal for characterizing hope: It is
doubtful that one's motivation always depends on the extent to which one hopes. I will argue (in section 4) that many
but not all realizations of hope enhance motivation. I conclude that we should not understand the thesis that hope is
irreducible as an ontological thesis about hope as a mental state. Rather, as I suggest in section 5, philosophers should
focus on the conceptual irreducibility thesis, meaning that the concept of hope is irreducible.

3 | AG A I N S T CO M P O U N D ACCO U NT S : B E LI E F A N D D E S I R E A R E N OT
N EC E S SA RY CO N D ITI O N S FO R H O PE

There is a deeply entrenched idea, which exerts a pull towards compound accounts: the idea that desire and belief
are required as necessary conditions for hope. Proponents of the standard definition and its developments rely on a
twofold, widely shared intuition: first, that we do not hope for what we take to be impossible (whereas we can wish
for what we take to be impossible), and second, that we do not hope for everything we take to be possible, but only
for outcomes we also take to be good, at least in some respect. Taking this as a starting point, the thesis that hope
consists of a desire for an outcome and a belief in that outcome's possibility (along with a third factor) seems plau-
sible. However, I suggest that we should find a different way of capturing what underlies the intuition about hope.
With regard to belief, it is doubtful that belief (even implicit belief) in the possibility of the object is required for
hope. We can call the position that hope requires, entails or is partly constituted by belief doxasticism about hope. On
doxastic theories of hope, S hopes that p only if S believes that p is possible. Nondoxasticism is the thesis that it is false
that S hopes that p only if S believes that p is possible.11 One argument for nondoxasticism is the argument from lin-
guistic data. It seems perfectly conceivable that a person could say “I hope that p, but I'm in doubt about whether p
is possible. Evidence for the possibility of p seems roughly on a par with evidence against the possibility of p. On these
grounds, I believe neither that p is possible nor that it is impossible”. This is a case of suspension of judgment about
possibility. If it is, in principle, possible to suspend judgment about q, which means not to believe q or not‐q, it is also
possible to suspend judgment about whether something is possible (where q = “p is possible”). In cases like this, it is
felicitous to attribute hope, despite the fact that the person is in doubt about the hoped‐for object's possibility. The
attraction of doxasticism could plausibly be accounted for by pointing to doxastic constraints on hope: A belief in the
impossibility of the object would destroy the person's hope.12 If that is true, then rather than entailing a belief in the
possibility of the object, hope would seem to entail the absence of a belief in its impossibility.13

11
I borrow these terms from the existing parallel debate about doxasticism/nondoxasticism with regard to faith (see e.g. Howard‐Snyder, 2019). The
following argument for nondoxasticism from linguistic data is also inspired by the analogous argument that Howard‐Snyder offers for nondoxasticism
about faith.
12
Note that this is a loose way of articulating the doxastic constraint on hope. One might ask what the relevant notion of impossibility is and whether
we should talk about belief, knowledge or something else. However, my point is that we can reformulate the belief component of the standard defini-
tion in terms of the absence of certain doxastic attitudes.
13
The idea that hope may only require the absence of certain beliefs has been mentioned in the debate, although the consequences I suggest here have
yet to be drawn. Pettit refers to the absence of belief in impossibility and belief in possibility in one and the same sentence: “I must not rule out the
hoped‐for possibility; I must believe that it could be realized” (Pettit, 2004, p. 153). Andrew Chignell explicitly acknowledges that “surely someone can
hope for p even if he or she has no actual beliefs at all […] about the modal status of p” (Chignell, 2014, p. 101).
6 | BLÖSER

One might object that the attraction of doxasticism cannot be taken care of merely in this negativist way, re-
quiring the absence of certain beliefs. One might be tempted to think that important hopes – in particular, action‐
guiding hopes or fundamental hopes that are constitutive of who we are – require a positive belief in possibility.
However, note first that it is plausible to assume that in many cases of hope, we also entertain the belief that the
outcome is possible. My point is merely that this is not necessary. There is no reason to assume that this is differ-
ent in the case of important hopes. A political activist might be in doubt as to whether it is really possible to realize
her ideal of social justice. Where the belief in realizability is absent, hope can be a stand‐in for this belief: We may
act solely on the basis of hope for social justice.
Acting on the basis of hope without belief in the possibility of the object might strike some as problematic
in cases where much is at stake and acting involves great risk. I agree, but this does not mean that hope, in such
cases, requires belief. Rather, it reflects the fact that we judge actions on the basis of mere hope in such cases to
be irrational; i.e. we tend to think that considerations of reasonable caution require the person to be more certain
of her success. Think of the two mountaineers: We might view Reinhold as irrational, in the sense of being too
willing to risk his life, if he climbs solely on the basis of hope without also having a justified belief in the possibility
of succeeding.
With regard to the relation between hope and desire, similar considerations apply. The true core of the stan-
dard definition is the following: we only hope for objects we see as good or attractive, at least in in some respect.
The standard definition suggests that desire (or similar attitudes, such as attachment)14 expresses this attitude
towards the hoped‐for outcome. However, there are two alternatives to this view: First, hope could be a pro‐atti-
tude sui generis. The category “pro‐attitudes” is often taken to comprise a variety of phenomena. Donald Davidson
lists, among others, “desires, wantings, urges, promptings, and a great variety of moral views, aesthetic principles”
(Davidson, 1963, p. 686), but also “emotions, sentiments, moods, motives, passions and hungers whose mention
may answer the question ‘Why did you do it?’” (Davidson, 1963, p. 689). Michael Smith, who takes up Davidson's
term “pro‐attitude”, considers desire a paradigmatic member of this class but concedes that there are other pro‐
attitudes and explicitly lists hope as one of them (Smith, 1987, p. 55). On this comprehensive view of pro‐attitudes,
hoping can be seen as a pro‐attitude in its own right, just like desiring, wishing, valuing, and the like. Indeed, if we
take Davidson's criterion for pro‐attitudes – mentioning them can answer the question “Why did you do it?” –
hope seems just as adequate as desire: The question “Why did you take part in the competition?” could be:
“Because I hoped to win!”.
This view explains not only why hope need not involve desire (because it is a pro‐attitude in its own right)
but also why we can sometimes substitute “hope” with “desire” or “wish”. As long as we mainly want to express
our pro‐attitude towards the outcome, all concepts can be adequate. It is only when differences between the
concepts matter that we cannot use them interchangeably. We can wish for what we take to be impossible, for
example, although we cannot hope for such outcomes. Thus we cannot substitute “wish” for “hope” in the sen-
tence “I wish I could fly”.
There is a second alternative to the standard view of seeing desire as a necessary condition for hope. One
might argue that desire and hope are different types of attitudes that do not necessarily occur together. The broad
category of pro‐attitudes, as Davidson conceives it, comprises very different attitudes, which arguably should be
distinguished. Most philosophers are interested in pro‐attitudes insofar as they dispose the subject to bring about
the state of affairs in question. Let me call this conception, which refers only to attitudes that aim to realize their
object, the “narrow” conception of a pro‐attitude. On a widespread account of desire, desires are simply disposi-
tions to act (e.g. Smith 1987).15 Hopes, by contrast, lack this essential link to action: They can be directed to past

14
This is the term preferred by Adrienne Martin insofar as she wants to highlight its nonrational nature (Martin, 2013, p. 7).
15
To be sure, there are other accounts of desire that hold that desires are evaluative representations, i.e. that they represent the desired object as
good in some way (e.g. Stampe 1987). Against the background of such a theory and the assumption that hope is likewise an evaluative representation
of the object as good, it is harder to make the argument that hope may exist without desire.
BLÖSER | 7

states of affairs or to future states of affairs for which the subject is conscious that they are not in his or her power.
I discuss this point in more detail in the next section. Here, I want to make the point that since a person can hope
without being disposed to bring about the hoped‐for outcome, hope is not a pro‐attitude in the narrow sense.
Thus, on a widespread theory of desire, this implies that hope can exist without desire.
In sum, a desire for the outcome and a belief in its possibility are not necessary conditions for hope because
it is possible to hope without entertaining the belief, and one can either view desire and hope as pro‐attitudes
in their own right or argue that hope and desire are different types of attitudes that do not necessarily occur to-
gether. As all compound accounts of hope contain the belief and desire condition, this amounts to a refutation of
all compound accounts.

4 | AG A I N S T H O PE A S A PR I M ITI V E M E NTA L S TATE: H O PE D O E S N OT

N EC E S S A R I LY H AV E M OTI VATI O N A L FO RC E

For Segal and Textor, hope is a primitive mental state that is characterized by its motivational role. In this section, I
want to argue against the idea that hope plays this role in all cases. Hence, we have no reason to assume that hope
is a primitive mental state that can be characterized by one single functional profile.
Remember Segal and Textor's formulation of hope's motivational role: “the extent of a subject's motivation for
doing A depends on the extent to which he hopes that if A, then G,” where “G is a possible desirable outcome of
an attempted action A” (Segal & Textor, 2015, p. 222). First of all, this characterization is not informative in cases
of hope for states of affairs that we cannot influence through our actions at all (e.g. hope for salvation, hopes for
the past). Further, the object of hope is typically not a conditional (“if A, then G”), but simply a state of affairs or
outcome (“G”).16 With regard to outcomes that we can influence with our actions, it might seem plausible that
hope contributes to motivation. However, we can have hopes without being engaged in realizing them – we might
hope for a just society and deliberately fail to do anything to promote it, and we might even hope to pass an exam
without doing our part to realize it, instead “just hoping” for good luck.
To be sure, one could try to rescue motivational force in dispositional terms (Segal & Textor, 2015, p. 213).
Surely, I can hope without acting on that hope at a given moment, perhaps due to lack of opportunity. Thus far,
nothing speaks against holding that this hope still has (dispositional) motivational force. However, imagine a case
in which I consistently treat my hope as a replacement for action. I tell myself: “I hope that we will find a solution
to climate change, but how that hope is realized is none of my business, so I won't forgo flying to far‐off holiday
destinations every year”. In such cases, it is meaningless to still ascribe motivational force in a dispositional way,
because it is uninformative to characterize the hope in question in terms of something that never becomes actual.
From the perspective of those who see motivational force as a crucial feature of hope, one might be tempted
to deny that such hopes are genuine instances of hope rather than mere wishes or fantasies. We should resist
this temptation, which is grounded in the wish to equalize all realizations of hope and to exclude those instances
that do not fit the preferred general characterization. To be sure, we might criticize hopes that fail to motivate as
worthless, and the hoping subject perhaps even as irrational. They are still genuine hopes, however, just like the
hopes that motivate agents (such as political activists) to perform actions.
In sum, hope is not necessarily motivating but can exist without (or even undermine) motivation. Hope is thus
not sufficient for being motivated – but is it perhaps necessary? We can imagine cases where this is true: It might
well be possible that without hope, Reinhold the mountaineer would not continue to climb. However, why should
that be true of all cases? Another mountaineer would perhaps be equally motivated in Reinhold's situation be-
cause he has willpower instead of hope. Hope might – at least in some cases – be an epiphenomenon with regard

16
To be sure, Segal and Textor's characterization might be plausible in cases of hope where the object is conditional (“I hope that if I do A, then G”),
but not many hopes have this conditional object.
8 | BLÖSER

to the motivation to act: Its presence or absence might be characteristic of the person's outlook in that situation,
without affecting her readiness to act.
Many authors generalize too quickly from one case of hope (or a class of cases) to all cases. When presenting the
contrasting cases that are used to criticize the standard definition in section 1, I mentioned that different authors use
counterexamples to point to different features that distinguish cases of hope from cases that do not involve hope,
e.g. having a positive outlook, motivational force, or seeing oneself as justified. On my approach, all of these inter-
pretations may be appropriate for the cases to which they refer. For example, Martin interprets the crucial difference
between Alan and Bess as a difference concerning what they take to be reasons: Alan does not see himself as being
justified in fantasizing about being healthy again, whereas Bess does. However, even though Martin may be right re-
garding this case – and probably other similar cases – it would be hasty to assume that this element must be present
in all cases of genuine hope. Rather, seeing oneself as justified in hopeful activities is one aspect of a specific class of
cases that fall under the concept of hope.
In the face of the many different realizations of hope, there is a heavy burden of proof to show that hope is one
and the same primitive mental state or amounts to one and the same compound of states in all cases. Instead, let
me now present the alternative that hope is multiply realizable by different mental states.

5 | H O PE A S A N I R R E D U C I B LE CO N C E P T

So far, I have argued that compound accounts of hope fail and that the thesis that hope is a primitive mental state
does not convince either. What option remains? In this section, I argue for a conceptual irreducibility thesis. This
means that hope is an irreducible concept (not analysable in terms of belief plus desire plus something else) and
can be realized in many ways (i.e. by different combinations of mental states). Following Wittgenstein, I suggest
that the various instances of hope are related by way of family resemblance.
Segal and Textor seem to regard it as obvious that we should understand the claim “Hope = Hope” (Segal
& Textor, 2015, p. 221) on an ontological level, meaning that hope is a mental state sui generis. However, hope
could be a concept that cannot be “defined away”, while still being realized by complex mental states. This would
mean that persons who hope may be in different mental states, where those cases all fall under the concept of
hope.
This suggestion does justice to the fact that we hope in a great variety of ways. The thesis that hope is an irreduc-
ible concept under which fall many different instances on the ontological level can be understood as a Wittgensteinian
account. In his Philosophical Investigations, Wittgenstein rejects definitions based on necessary and sufficient con-
ditions, instead relying on the idea of “family resemblances” (Wittgenstein 1968 [1953], §67) to describe how very
different particular cases may fall under the same concept. Wittgenstein famously expounds this idea with regard
to games:

Consider for example the proceeding we call ‘games’. I mean board‐games, card‐games, ball‐games,
Olympic games, and so on. What is common to them all? – Don’t say: ‘There must be something in
common, or they would not be called “games”’ – but look and see whether there is anything common
to all. – For if you look at them you will not see something that is common to all, but similarities,
relationships, and a whole series of them at that. (Wittgenstein 1968 [1953], §66)

He continues:

I can think of no better expression to characterize these similarities than ‘family resemblances’;
for the various resemblances between members of a family: build, features, colour of eyes, gait,
temperament, etc. etc. overlap and criss‐cross in the same way. (Wittgenstein 1968 [1953], §67)
BLÖSER | 9

Just as there is no common core of all games, there seems to be no common core of all hopes. One might object
that it seems clear that there are characteristics common to all games: They are all activities and take place at a cer-
tain time, etc. What Wittgenstein should have said is “you will not see something common to all games in virtue of
which they are called games” (Griffin, 1974, p. 637). Similarly, it is plausible to say that hope involves some (mental)
activity. However, this condition is so general that it does not help to define hope or distinguish it from other mental
activities.
A thorough defence of hope as a family resemblance concept, which I cannot deliver here, would highlight
negative as well as positive aspects of family resemblance: In a positive sense, the idea is that we can identify cer-
tain interesting features that some, though not all, cases share (although, strictly speaking, Wittgenstein's account
seems to allow for instances which are similar to the others without sharing a single feature). I focused here on the
negative sense, i.e. the claim is that there is no common core to all instances. Compound accounts see this com-
mon core in a set of necessary and sufficient conditions. Against this view, I argued that belief and desire are not
necessary conditions. Segal and Textor's view that hope is a simple mental state similarly assumes such a common
core, not in terms of conditions but rather in the form of a mental state sui generis. A Wittgensteinian account goes
against both ways of conceiving of a common core of hope.

6 | CO N C LU S I O N

I have argued that the concept of hope is irreducible to belief and desire. I surveyed criticisms of developments of
the standard definition that show that existing compound accounts face problems. However, the only alternative
– Segal and Textor's thesis that hope is a primitive mental state – cannot convince either. First, it does not robustly
refute compound accounts. Second, it overestimates the motivational role of hope. In order to remedy the first
problem, I argued that the concepts of belief and desire are not even necessary conditions for hope. No positive
belief in the possibility of the outcome is required – only the absence of a belief in its impossibility. Rather than
assuming that hope involves desire, we should acknowledge either that the concept of hope designates a pro‐at-
titude sui generis or that desire is a pro‐attitude in the narrow sense while hope is not – thus, hope and desire can
come apart. As all compound accounts are based on the standard account in taking desire and belief to be neces-
sary components, my argument amounts to a refutation of all compound accounts of hope that are based on de-
sire and belief.17 Discussion of the second problem shows that hope can play different motivational roles.
Therefore, I conclude that persons who hope do not necessarily share one and the same primitive mental state
that is characterized by motivational force. One promising alternative is to understand hope as an irreducible
concept that is multiply realizable on an ontological level. The various ways in which we hope are related by way
of family resemblance.

AC K N OW L E D G E M E N T S

I want to thank Andrew Chignell, Ilaria Cozzaglio, Jakob Huber, Sofie Moeller, Darrel Moellendorf, Andi Müller,
Elke Elisabeth Schmidt, Matthias Schmitt, Markus Schrenk, Marcus Willaschek and the audience of the conference
“New Directions in the Philosophy of Hope” in Frankfurt for very helpful discussions and comments on earlier
versions of this text.

17
Note that in principle, there might be compound accounts that understand hope as having necessary conditions other than desire and belief. The
only possibility I am aware of is the view that hope is an emotion, where an emotion is understood as a representational, felt evaluation. Sabine Döring
describes hope as an emotion that represents the object as good and “as possible, insofar as we feel confident that this good will be brought about”
(Döring, 2014, p. 126). For reasons of space, I cannot discuss this view here. Denying that belief and desire are necessary conditions for hope is com-
patible with accepting the view that hope represents the object as good and possible. However, I am sceptical of the claim that hope always involves
a feeling of confidence, especially if we hope for something that is very improbable.
10 | BLÖSER

ORCID

Claudia Blöser https://orcid.org/0000-0001-7139-9686

REFERENCES

Bovens, L. (1999). The value of hope. Philosophy and Phenomenological Research, 59, 667–681.
Calhoun, C. (2018). Doing valuable time: The present, the future, and meaningful living. Oxford, UK: Oxford University Press.
Chignell, A. (2014). Rational hope, possibility, and divine action. In G. Michalson (Ed.), Kant’s religion within the boundaries
of mere reason: A critical guide (pp. 98–117). Cambridge, UK: Cambridge University Press.
Davidson, D. (1963). Actions, reasons, and causes. Journal of Philosophy, 60(23), 685–700.
Döring, S. (2014). What may I hope? Why it can be rational to rely on one's hope. European Journal for Philosophy of
Religion, 6(3), 117–129.
Downie, R. S. (1963). Hope. Philosophy and Phenomenological Research, 24, 248–251.
Griffin, N. (1974). Wittgenstein, universals, and family resemblances. Canadian Journal of Philosophy, 3, 635–651.
Han‐Pile, B. (2017). Hope, powerlessness, and agency. Midwest Studies in Philosophy, XLI, 175–201.
Howard‐Snyder, D. (2019). Three arguments to think that faith does not entail belief. Pacific Philosophical Quarterly, 100,
114–128.
Kwong, J. (2019). What is hope? European Journal of Philosophy, 27, 243–254.
Martin, A. M. (2013). How we hope: A moral psychology. Princeton, NJ: Princeton University Press.
Meirav, A. (2009). The nature of hope. Ratio, 22, 216–233.
Milona, M. (2018). Finding hope. Canadian Journal of Philosophy.
Milona, M., & Stockdale, K. (2018). A perceptual theory of hope. Ergo: An Open Access Journal of Philosophy, 5, 203–222.
Moellendorf, D. (2006). Hope as a political virtue. Philosophical Papers, 35, 413–433.
Pettit, P. (2004). Hope and its place in mind. Annals of the American Academy of Political and Social Science, 592, 152–165.
Segal, G., & Textor, M. (2015). Hope as a primitive mental state. Ratio, 28, 207–222.
Smith, M. (1987). The humean theory of motivation. Mind, 96(381), 36–61.
Stampe, D. W. (1987). The authority of desire. Philosophical Review, 96, 335–381.
Wittgenstein, L. (1968 [1953]). Philosophical investigations (G.E.M. Anscombe, Trans.). Oxford, UK: Basil Blackwell & Mott.

How to cite this article: Blöser C. Hope as an irreducible concept. Ratio. 2019;00:1–10. https​://doi.
org/10.1111/rati.12236​