Delusions, beliefs, and actions: Pragmatic commitment reconsidered

1
Delusions, beliefs, and actionsPragmatic
commitment reconsidered
Fabio Paglieri, Ryan McKay
European Society for Philosophy and Psychology -
Belfast, 25 August 2006
2
Motivation

• To analyze the main objection raised against the
  idea that delusions are beliefs the lack of
  pragmatic commitment argument (LoPC)
• To criticize the LoPC argument, using notions
  derived from epistemology and action theory
• To reformulate the LoPC argument
• To assess the import of this reformulation for a
  theory of delusions

3
What is a delusion?

• Informal definition A false mental
  representation held with strong conviction
  despite manifest evidence to the contrary
• Delusions range from the exotic (e.g. Im God)
  to the relatively mundane (My neighbours are
  persecuting me)
• Delusions are first-rank symptoms of
  schizophrenia and feature prominently in numerous
  other psychiatric and neurological conditions

4
Examples of delusion

• Examples of monothematic delusions
• Capgras an impostor substituted for a spouse or
  close relative
• Cotard being dead, non-existent, putrefying or
  missing internal organs
• Fregoli being constantly followed by someone in
  disguise
• Many others
• Polythematic delusions figure often in cases of
  schizophrenia (e.g. John Nashs fantasies)

5
The standard view Delusions as beliefs

• Clinical definition Delusion A false belief
  based on incorrect inference about external
  reality that is firmly sustained despite what
  almost everyone else believes and despite what
  constitutes incontrovertible and obvious proof or
  evidence to the contrary (DSM-IV-TR, 2000 821)
• Defenders of the doxastic view agree with the DSM
  that delusions are beliefs, but may disagree on
  other points e.g., falsity (Peters, 2001)

6
Objections to the doxastic view

• Lack of pragmatic commitment Deluded individuals
  frequently fail to act in accordance with their
  delusions
• Lack of affective commitment The emotive and
  affective responses seem inconsistent with the
  contents of their delusions
• Inability to provide convincing reasons
  Difficulties in providing reasons for their
  delusions, either because impossible to find
  (e.g. the world will end tomorrow), or because
  they seem not to be operative
• Lack of integration with other beliefs Often
  unconcerned by patent contradictions between
  their delusions and their other beliefs

7
Alternatives to the doxastic view

• Meta-cognitive thesis
• Currie, 2000 Currie, Jureidini, 2001 Currie,
  Ravenscroft, 2002 Stephens, Graham, 2004
• No-content thesis
• Jaspers, 1963 Berrios, 1991 1996
• Loss-of-understanding thesis
• Campbell, 2001 Eilan, 2001

8
The LoPC argument

• A given behaviour of a deluded patient
  contradicts or does not follow from the content
  of his/her delusion
• Therefore, he/she does not really believe the
  content of that delusion
• Example Many Capgras patients do not file a
  missing person report to the police, in spite of
  their delusion that the spouse has been
  substituted by an impostor. Hence, they do not
  truly believe in their delusion
• Bel(p) ? Do(A) therefore not-Do(A) ? not-Bel(p)

9
Key premises of the LoPC argument

• Pragmatic commitment is lacking in delusional
  patients
• challenged by Bayne and Pacherie (2005), since
  the evidence is not so B/W in this respect
  (ODwyer, 1990 Förstl et al., 1991 Young,
  Leafhead, 1996 Young, 2000)
• A certain behaviour or action can be used to
  pinpoint a single, specific mental attitude
• The mental attitude revealed by action is belief

10
Acting in spite of ones beliefs

• Observation Sometimes it seems rational to act
  in spite of our beliefs, i.e. either accepting
  for practical purposes something that we do not
  believe, or failing to act on something we do
  believe
• Hypothesis It may be convenient to split two
  functions that were usually collapsed in belief,
  i.e. aiming at being true, and being used in
  shaping action
• Belief A mental representation that the subject
  considers to be true, i.e. a correct image of
  reality
• Acceptance A mental representation that is used
  to plan and carry on an action, as a map by which
  the subject steers (Ramsey, 1931)

11
Varieties of Acceptance (Bratman, 1992)

• When do acceptance and belief diverge?
• Simplification of ones reasoning
• e.g. taking for granted it will not rain in
  planning a summer day out
• Asymmetries in the cost of errors (prudential
  reasons)
• e.g. acting on pessimistic presumptions to avoid
  costly surprises
• Needs of social cooperation
• e.g. acting on socially agreed premises (possibly
  without believing them individually)
• Special relations to others
• e.g. acting as if my close friend is innocent, in
  spite of strong evidence to the contrary
• Pre-conditions for any practical reasoning at all
• e.g. a soldier in a war zone planning his day as
  if he was confident he will survive it, although
  he is rather sceptical about this

12
Acceptance for prudential reason

• I plan for a major construction project to begin
  next month. I need to decide now whether to do
  the entire project at once or instead to break
  the project into two parts, to be executed
  separately. The rationale for the second strategy
  is that I am unsure whether I presently have the
  financial resources to do the whole thing at
  once. I know that in the case of each
  sub-contractor carpenter, plumber, and so on
  it is only possible at present to get an estimate
  of the range of potential costs. In the face of
  this uncertainty I proceed in a cautious way In
  the case of each sub-contractor I take it for
  granted that the total costs will be at the top
  of the estimated range. On the basis of these
  assumptions I determine whether I have at present
  enough money to do the whole project at once. In
  contrast, if you offered me a bet on the actual
  total cost of the project the winner being the
  person whose guess is closer to the actual total
  I would reason differently. (Bratman,
  1992 6)

13
Import for the LoPC argument

• This distinction defuses the LoPC argument,
  because
• Actions reveal acceptances, not beliefs
• Beliefs and acceptances do not always coincide
• Example Failure to file a missing person report
  does not imply that the Capgras patient does not
  believe the delusional content. Instead, this
  content could be truly believed, and yet not
  acted upon due to pragmatic concerns (e.g.
  prudential reasons)
• The belief vs. acceptance distinction undermines
  the 3rd premise of the LoPC argument Delusions
  could be beliefs that are not accepted for
  practical purposes

14
Critique of the B/A distinction

• The distinction is merely stipulative Actions
  reveal acceptances rather than beliefs just
  because we have defined these notions accordingly
• The distinction turns belief into an unverifiable
  quantity Since beliefs are not revealed by
  action, and verbal behaviour is just one kind of
  action, then how can we attribute beliefs
  meaningfully to anyone?
• Legitimate doubt Do we really need the notion of
  acceptance, in order to account for those cases
  in which we seem to act in spite of our beliefs?

15
Alternative explanation of Bratmans example

• The subject believes that
• The project costs are likely to be in the average
  of the expected range
• However, it is possible, albeit unlikely, that
  they will escalate to the top of the range
• If the latter happens, the subject will be in
  deep financial troubles
• The subject intends to
• Conclude the construction works in a reasonable
  amount of time
• Avoid the risk of bankruptcy
• By bringing into the picture beliefs about
  possibilities and motivations (and why not?),
  this case is explained without any need to
  contrive an additional notion of acceptance as
  independent from belief

16
Two birds with one stone?

• This critique undermines the B/A distinction, but
  at the same time
• calls into question the 2nd premise of the LoPC
  argument, i.e. the idea that a certain action is
  indicative of a single mental attitude
• The B/A distinction pointed at the right problem
  in the LoPC argument (the complex nexus between
  belief and action), but offered the wrong
  solution (dissociating belief from action)

17
Critique of the LoPC argument

• Our alternative explanation of the construction
  project example shows that a certain behaviour
  cannot be used as a litmus test for a single
  belief, but rather reveals a complex system of
  beliefs and motivations
• The same applies to delusional cases The Capgras
  patient who fails to file a missing person report
  may have excellent reasons ( other beliefs
  motivations) for doing so, and if we have grounds
  for attributing him/her such reasons, the LoPC
  objection does not apply
• The LoPC argument must be reformulated, to avoid
  over-simplification of the link between actions,
  beliefs, and motives

18
LoPC argument reconsidered

• Given a single, specific delusional content, we
  take any instance of behaviour that cannot be
  properly explained by a system of beliefs (plus
  motivations) which includes the delusional
  content as indicating that the subject is not
  acting on the ground of his/her delusion,
  therefore he/she does not believe it.
• Sensible concern Considering that there are
  potentially infinite stories that we can devise
  to account for ones action, is this
  reformulation overly prudent or conservative? Are
  we making the LoPC objection too weak?

19
LoPC revised Key constraints

• Justification Additional beliefs and motives
  must be backed up by some evidence in the
  clinical history of the patient
• No missing person report We cannot attribute to
  a Capgras patient the belief that his wife ran
  away of her own accord, unless we have evidence
  to that effect
• Cross-behavioural consistency They must be
  consistent with beliefs and motives invoked to
  explain other behaviours of the patient
  (controversial, here meant as a practical
  guideline for the observer and not as a cognitive
  law for the subject)
• No missing person report We cannot attribute to
  a Capgras patient the goal of concealing his
  bizarre conviction from others, if he is known to
  frequently declare it in public places

20
Musings on behavioural incongruence

• Pragmatic commitment in delusional patients is
  not so much lacking, as rather inconsistent
• Cases of non-verbal delusion-based behaviour
• Verbal delusion-based behaviour, i.e. patients
  frequently assert the content of their delusions
• From behavioural inertia to behavioural
  incongruence
• Our reformulation of the LoPC argument provides a
  framework for dealing with behavioural
  incongruence The contextual differences in the
  pragmatic commitment showed towards ones
  delusions may be attributed to other relevant
  beliefs and motives of the deluded subject

21
Musings on behavioural incongruence

• Why the diversity of PC across different
  non-verbal behaviours?
• HP More delusion-based action in presence of
  positive expectations on the outcome
• Comparing Capgras delusion (Young, 2000) and
  erotomania (ODwyer, 1990)
• Fact Delusion-based action is more frequent in
  the second condition
• Speculation Erotomaniacs are usually driven by
  positive expectations, even when they act
  violently (e.g. John Hinckley Jr. shooting Regan
  for the love of Jodie Foster) the same does not
  necessarily apply to Capgras patients
• Limit Heterogeneous data, never meant to be
  compared

22
Musings on behavioural incongruence

• Why PC is much more frequent in verbal than in
  non-verbal behaviour?
• Talk is cheap, or cheaper than other actions
• Increase in delusion-based non-verbal behaviour
  towards trusted interlocutors
• Anecdotal evidence (thanks to Maria Miceli)
• An old lady, paranoid in several respects,
  developed the conviction that the staff at her
  nursing home wanted to kill her. This lady is
  perfectly capable of voicing her belief and even
  providing detailed reasons of why she formed it,
  and she does so with trusted people, e.g. close
  friends and some of her fellow patients. However,
  she sternly refuses to give away this information
  to the staff of the nursing home for the
  obvious reason, she says, that this would make
  them realize that she is aware of their
  conspiracy, and this would further jeopardize her
  predicament.

23
Conclusions and future work

• The standard formulation of the LoPC argument is
  mistaken, not because action does not relate to
  belief, but because this relation is always
  mediated by other beliefs and motives
• It is possible to reformulate the LoPC argument,
  rather than abandoning it
• This reformulation will have to be tested against
  many cases of delusions, to check if it is
  applicable, and what are its consequences for the
  debate about the nature of delusions
• This reformulation also provides a framework for
  discussing behavioural incongruence in delusional
  patients
• These speculations await further empirical
  verification

24
Thanks for your kind attention.
Fabio Paglieri paglieri_at_media.unisi.it Ryan
McKay rmckay_at_csu.edu.au
European Society for Philosophy and Psychology -
Belfast, 25 August 2006
25
Outline

• Definition and examples of delusions
• The doxastic view Delusions as beliefs
• Main criticisms against the doxastic view
• The LoPC argument
• 1st attack Belief vs. acceptance
• Critique of the distinction between belief and
  acceptance
• 2nd attack Single beliefs vs. systems of beliefs
  and goals
• The LoPC argument reconsidered
• Some thoughts on behavioural incongruence
• Conclusions and future work

26
Alternatives to the doxastic view

• Meta-cognitive thesis Delusions as higher order
  cognitive phenomena, i.e. meta-evaluations.
  Delusions as imaginings, which the subject
  mistakenly consider to be believed
  (meta-beliefs). Delusions as cognitive
  hallucinations (Currie, 2000 Currie, Jureidini,
  2001 Currie, Ravenscroft, 2002), revealing a
  complex delusional stance (Stephens, Graham,
  2004).

27
Alternatives to the doxastic view

• No-content thesis Certain delusions are so
  bizarre (violating biological, physical and
  logical laws) that they are devoid of any
  meaningful content (Jaspers, 1963 Berrios,
  1991). Delusions are merely empty speech-acts
  that disguise themselves as beliefs (Berrios,
  1996 126).

28
Alternatives to the doxastic view

• Loss-of-understanding thesis The way that
  deluded subjects articulate their delusions
  connotes a loss of understanding of the socially
  shared meaning of the expressions that they are
  using (Campbell, 2001 Eilan, 2001). This implies
  that the delusional content (believed or not)
  differs markedly from what would be implied in
  normal usage of the same propositions.

29
Defining Features Acceptance vs. belief
30
B/A as Collective Mental States

• The problem of collective doxastic states
• Is it possible to attribute mental states such as
  beliefs and acceptances to collective entities
  (e.g., groups, institutions, etc.) in a
  non-summative sense?
• If so, are these collective mental states to be
  interpreted as either beliefs or acceptances?
• Disagreement believers (Gilbert, Tollefsen) vs.
  rejectionists (Wray, Meijers)
• Distinction between two sub-types of acceptance
  (Engel, Tuomela)
• acceptance as true / assent vs. pragmatic
  acceptance

31
Pragmatic Acceptance vs. Acceptance as True
32
An Alternative Definitional Strategy

• Instead of listing similarities / differences, we
  try to
• find a truly distinctive feature
• show how all other differences follow from such
  trait
• Hakli (in press)
• voluntary acceptance vs. involuntary belief
  (criticized by Bratman and Tollefsen)
• Our proposal
• pragmatic value (acceptance)vs. truth-functional
  value (belief)

33
Belief and Acceptance in a Nutshell

• Bel(p) is usually a good reason to Acc(p), but
  there are exceptions(cf. Bratman)
• Acc(p) may lead, if the resulting action is
  successful, to Bel(p)(cf. Pascals argument)

34
Belief vs. Acceptance in Formal Models

• Bayesian models distinction not mentioned
• Reason Maintenance Systems distinction not
  mentioned
• Doxastic logics distinction not mentioned or
  collapsed
• e.g. in BDI beliefs are both truthful
  representations of external reality and a basis
  for deliberation and action
• AGM belief revision distinction shortly
  mentioned according to Stalnakers inclusive view
• Acceptance is a broader concept than belief it
  also includes such attitudes as assuming,
  presupposing, and positing. To accept a
  proposition is to treat it as true in one way or
  another. This position entails that
  acceptance is relative to context A person may
  accept something in one context but reject it or
  leave it indetermined in another. On my
  account acceptance entails full belief.
  (Gärdenfors, 1988 23)

35
Beliefs vs. Acceptances Does It Matter?

• Beliefs vs. acceptances
• Significant differences between reasons to
  believe and reasons to accept An
  information-based approach may accommodate both
• Different properties that should be considered to
  avoid confusion / ambiguity in formal models (see
  Cohens accusation)

36
Two Puzzles and One Doubt

• There seems to be valid reasons to improve the
  epistemological foundations of formal models of
  belief dynamics (Pollock, Gillies, 2000). But
  this leaves open two questions and one doubt
• Question of Primacy What does it come first (in
  both ontogenesis and phylogenesis), belief or
  acceptance?
• Question of Function If belief is a derivative
  notion, why is it developed at all? What is the
  adaptive function of belief thus conceived? Are
  we insanely obsessed with truth, or is there an
  added value in belief dynamics?
• and yet
• Radical doubt Is acceptance for real, or is it
  just a contrivance of harm-chair philosophy?
  Would it be possible to reduce alleged examples
  of acceptances to innocuous cases of belief?