International Journal of Law and Psychiatry xxx (2014) xxx–xxx

IJLP-01044; No of Pages 8

Contents lists available at ScienceDirect

International Journal of Law and Psychiatry

Mental disorder and legal responsibility: The relevance of stages of decision making

Annemarie Kalis a,⁎, Gerben Meynen b

a Department of Philosophy, Utrecht University, Janskerkhof 13A, 3512 BL Utrecht, The Netherlandsb Tilburg Law School, Department of Criminal Law, Tilburg University, P.O. Box 90153, 5000 LE Tilburg, The Netherlands

⁎ Corresponding author. Tel.: +31 30 253 4399.E-mail addresses: a.kalis@uu.nl (A. Kalis), g.meynen@u

1 In this paper we focus on psychiatrists, but much of wis also relevant for psychologists.

2 The question what constitutes a ‘substantial role’ is alegal frameworks, more details on this below.

3 See, e.g., Elliott (1996), Robinson (1998), and Simon anare, for instance, reflected in the M'Naghten Rule, C is reflTest/Rule, while A and C are reflected in the Model Penal

http://dx.doi.org/10.1016/j.ijlp.2014.02.0340160-2527/© 2014 Elsevier Ltd. All rights reserved.

Please cite this article as: Kalis, A., &Meynen,Journal of Law and Psychiatry (2014), http://

a b s t r a c t

a r t i c l e i n f o

Available online xxxx

Keywords:Legal responsibilityDecision makingInsanity defenseMental disorder

The paper discusses the relevance of decision-making models for evaluating the impact of mental disorder onlegal responsibility. A three-stage model is presented that analyzes decision making in terms of behavioralcontrol. We argue that understanding dysfunctions in each of the three stages of decision making could provideimportant insights in the relation betweenmental disorder and legal responsibility. In particular, it is argued thatgenerating options for action constitutes an important but largely ignored stage of the decision-making process,and that dysfunctions in this early stage might undermine the whole process of making decisions (and thus be-havioral control) more strongly than dysfunctions in later stages. Lastly, we show how the presented frameworkcould be relevant to the actual psychiatric assessment of a defendant's decision making within the context of aninsanity defense.

© 2014 Elsevier Ltd. All rights reserved.

1. Introduction

In court, forensic psychiatrists and psychologists1 are regularly askedto assess a defendant's mental conditionwithin the context of an insan-ity defense (Rogers & Shuman, 2005). If they decide that psychiatricillness played a substantial role2 in the commitment of the crime, thecourt may rule that the defendant does not bear legal responsibilityfor their action. There is ample literature on considerations that couldinform a psychiatric assessment within the context of an insanitydefense, which partly depend on the various legal rules that guidesuch a defense, such as the M'Naghten Rule, Durham Rule, and ModelPenal Code (Borum & Fulero, 1999; Buchanan, 2000; Elliott, 1996;Morse, 2011; Rogers & Shuman, 2005). In general, criteria that oftenfeature in legal insanity standards are:

A) that defendants did not know their behavior was morally wrongB) that defendants did not know what they were doingC) that defendants had no control over what they were doing.3

Although differences between legal insanity standards might par-tially reflect different views on the general conditions for responsibility,

vt.nl (G. Meynen).hat we state about psychiatrists

nswered differently in different

d Ahn-Redding (2006). A and Bected in the Irresistible ImpulseCode.

G.,Mental disorder and legal rdx.doi.org/10.1016/j.ijlp.2014

they could also be seen as different ways of expressing the same under-lying idea: a mental disorder may exculpate defendants if the disordercompromised their ability to make decisions for action.4 After all, thedifferent components mentioned appear to be relevant as far as theyhave influenced a defendant's decision making about the course of ac-tion (Meynen, 2013). The idea that mental disorders may underminea person's responsibility for actions by undermining decision-makingprocesses is not only reflected in legal regulations, but also in moralphilosophy (Kalis, 2011; Meynen, 2010;Wallace, 1994). In fact, the ele-ments of a ‘healthy’ decision-making process could be seen as elementsof normal behavioral control by an agent. In standard situations, wehold each other morally responsible for those actions that are self-initiated and the result of a decision-making process in which the sub-ject could, if necessary, consciously intervene. We therefore do nothold people morally responsible for behaviors like sneezing or reflexmovements, or behaviors brought about by coercion (Kalis, 2011;Wallace, 1994). However, in the context of this paper we are primarilyconcerned with legal responsibility, and it must be noted that moraland legal responsibility do not always go together. We take moral re-sponsibility to be a prerequisite for legal responsibility (for an overviewof different responsibility concepts see Vincent, 2011). In the presentpaper, we will consider legal insanity as a condition in which thedefendant's decision-making process was dysfunctional to such an ex-tent that the defendant no longer bears legal responsibility for the be-havior resulting from it.

4 As Alec Buchanan writes: “If psychiatric conditions are to be grounds for exculpation,they must impair the sufferer's ability to choose. There are many ways in which they maydo this” (Buchanan, 2000, p.80). In discussing decisionmaking in this paper,we only focuson making decisions for action, and thus leave aside other types of decision making.

esponsibility: The relevance of stages of decisionmaking, International.02.034

5 An extended version of this comparative analysis of stagemodels can be found inKaliset al. (2008). On option generation see also Kalis, Kaiser, and Mojzisch (2013).

2 A. Kalis, G. Meynen / International Journal of Law and Psychiatry xxx (2014) xxx–xxx

Althoughmuch has been written about issues related to legal insan-ity, central questions in the field of criminal law and forensic psychiatryand psychology remain: how to optimally assess a defendant's decision-making capacities at the time of the crime, and how to standardize judg-ment procedures to such an extent that equality before the law is guar-anteed, while also leaving room for those aspects of clinical judgmentand experience that may resist rigid standardization? In this paper weaim to contribute to these debates, proposing an underlying theoreticalframework derived from (neuro)psychological and philosophical re-search on decision making, that could contribute to the specificationand conceptual justification of criteria for criminal responsibility asthey are used in different legal systems. We will not focus on specificcriteria employed in individual jurisdictions (Meynen, 2012, 2013).More precisely, we aim to transcend the boundaries of particular juris-dictions in order to develop a general and conceptual perspective onmental disorder and legal responsibility. For instance, when we usethe phrase ‘a disorder compromised one's ability to make decisions foraction’ (see above), this is a conceptual statement rather than theword-ing of a particular legal insanity standard. In fact, even if a legal insanitystandard does not explicitly mention ‘compromised decision making’,this does not rule out the relevance of decision making to the phenom-ena mentioned in these standards (see also above). For instance, acapacity for decision making appears to be implied in the Model PenalCode when it states that “…a person is not responsible for criminalconduct if at the time of such conduct as a result of mental disease ordefect he lacks substantial capacity either to appreciate the criminalityof his conduct or to conform his conduct to the requirements of thelaw” (American Law Institute, 1962). This statement implicitly refersto decisionmaking, as we typically conform our conduct to the require-ments of the law by making decisions about our conduct (see alsoMeynen, 2013).

The M'Naghten Rule focuses on defendants' knowledge, instead ofon their conduct: “…the party accused was labouring under such a de-fect of reason, from disease of the mind, as not to know the nature andquality of the act he was doing; or, if he did know it, that he did notknow he was doing what was wrong.” (M'Naghten's Case, 10 Cl. & Fin.200, 8 Eng. Rep. 718, H.L. 1843). The link between knowledge and deci-sion making may appear to be weaker than between conduct and deci-sion making. Still, the idea underlying the M'Naghten standard appearsto be that compromised insight in the nature, quality, or wrongfulnessof the actmay lead to compromised decisions about the course of action(see alsoMeynen, 2013, and Section 4). Clearly, it is beyond the scope ofthis article to discuss all legal insanity standards and to determine howeach of them relates to the idea of ‘compromised decision making’. It iseven possible that not everybody agrees that a particular legal standardreflects compromised decision making on the part of the defendant.Still, we assume that most will agree that in general, the impact of amental disorder on a defendant's decision making at the moment ofthe crime is relevant to psychiatric assessments within the context ofan insanity defense (see also Meynen, 2013).

The proposal developed in this paper consists of three steps. First,weoutline a stage model that distinguishes between three aspects of deci-sion making, as suggested by Kalis, Mojzisch, Schweizer, and Kaiser(2008). Second, we discuss the impact that impairments in each ofthese stages could have on a person's actions. Third, we argue thatimpairments in different stages might have different implications forlegal responsibility. Therefore, this framework could inform actualassessments of defendants as well as research on decision making inpersons suffering from mental disorder.

2. A three-stage model of decision making

The proposal we aim to develop in this paper is that in assessing theimpact of psychiatric dysfunctions on decisionmaking, it is important todistinguish three different stages of the decision-making process, and toinvestigate how different dysfunctions can affect each of these stages

Please cite this article as: Kalis, A., &Meynen, G.,Mental disorder and legal rJournal of Law and Psychiatry (2014), http://dx.doi.org/10.1016/j.ijlp.2014

(Kalis et al., 2008). The framework divides the decision-making processin the stages: option generation, option selection, and action initiation(Fig. 1). These stages should capture decision making in general, al-though we do not claim that every decision-making process proceedsin a strict linear fashion: feedback loops, for instance, are likely tooccur. The main aim of the model is to distinguish different and crucialelements or aspects of the decision-making process, which are, to alarge extent, dependent on one another.

This framework, introduced in Kalis et al. (2008), builds upon andexpands existing sequential models of decision making and action(Ernst & Paulus, 2005; Heckhausen & Gollwitzer, 1987; Heckhausen &Heckhausen, 2008). For example, the Rubicon model developed byHeckhausen and Gollwitzer consists of four action phases. Firstly, inthe predecisional phase (1) different options for action are evaluatedin terms of their desirability and feasibility. When this evaluative pro-cess leads to a decision, one moves on to the postdecisional phase (2).In this phase the focus is on transforming a decision into action, thuson planning. As soon as the person takes steps toward actual executionof the action, the process moves on to the actional phase (3). After theaction has been performed, the action is evaluated: this is referred toas the post-actional phase (4). According to Heckhausen (Heckhausen& Gollwitzer, 1987; Heckhausen & Heckhausen, 2008), information pro-cessing during both the predecisional phase and the postactional phaseis open-minded and impartial. However, during the postdecisionalphase and the actional phase, information processing is thought to be bi-ased in favor of the chosen alternative. A similar butmore concise stage-model has more recently been developed by Ernst and Paulus (2005),who distinguish three phases: (1) formation and evaluation of prefer-ences regarding different options, (2) selection and execution of theaction; and (3) action evaluation.5

When we compare our model to these earlier models, some differ-ences should be noted. First, we focus on the stages up to, and including,the initiation of the action, which means that we exclude the stage ofoutcome evaluation. This is because the legal investigation in the contextof an insanity defense mostly focuses on the defendant's decision-making process up to the unlawful act. More importantly, we includethe stage of generating options for actions as a separate stage in ourmodel. Option generation has so far been largely ignored in decision-making research; we argue that this aspect of decision-making is partic-ularly relevant in addressingquestions of legal insanity. Thirdly, contraryto most existing models our stages refer to transition points in thedecision-making process: for example, we use the term optionselection to refer to the point where an actual decision is made.For Heckhausen, this point would lie ‘in between’ the predecisionaland the post-decisional phase. In our model, option generation refersto the transition point between a phase where no options are availableand the phase that Heckhausen would describe as the predecisionalphase. In the remainder of this section, we briefly discuss each of thethree stages included in ourmodel in the normal (nonpathological) con-dition. In the next section,we relate the three stages to psychopathology.

Stage 1 In order to decide what to do in a certain situation, one mustcome to see that certain options are available for action. Werefer to this stage as option generation. One can come up withoptions for action via very different mental operations, for ex-ample, bymemory retrieval, by creative processes, or by direct-ly perceiving possibilities in one's environment (Kalis et al.,2008; Smaldino & Richerson, 2012). Meanwhile, what thesedifferent processes have in common is that they determineour range of possibilities in a certain situation: specific coursesof action that are open to us. Our behavioral repertoire is broad,but always limited by the actual options we generate. ‘Options’could be defined as representations of candidates for goal-

esponsibility: The relevance of stages of decisionmaking, International.02.034

Fig. 1. A stage model of decision making (see Kalis et al., 2008).

6 We use these words in their everyday meaning. We do not intend to suggest that allaspects of our decision making need to be conscious. See Morse (2003) on (ir)rationalityas related to legal insanity.

3A. Kalis, G. Meynen / International Journal of Law and Psychiatry xxx (2014) xxx–xxx

directed action. Each generated option comes with a certainlevel of attractiveness or evaluation, although this evaluationis certainly not rigid or fixed — adjustment on the basis of re-flection or feedback is possible. For instance, one may considerwhat to do tonight. The options that may arise are: go to thecinema, work, visit a friend, clean the house, or watch TV.These are the options our mind generates, and we can pickone of them. Now, during the whole process of deciding whatto do, new options may be generated, but without optionsbeing generated, there are nodecisions to bemade. The optionsprovide thenecessary, albeit raw,material for all actual decisionmaking. They are, in fact, important enablers of the secondphase: the stage of option selection.

Stage 2 Option selection concerns both the further evaluation ofoptions and the actual selection of one of the options: decidingbetween the different candidates that are taken into account.As said, it is possible that the process of evaluation causesnew options to become available or apparent, which impliesa feedback loop to the stage of option generation, and thesewill then have to be taken into account as well. In any case,this is the phase of deliberation, of weighing pros and consand finally picking one of the options. The result of pickingone of the generated options, our ‘choice’, however, is not theendpoint of making decisions according to our model. A thirdphase is needed, action initiation.

Stage 3 Action initiation represents those processes that transform thechosen option into actual behavior. The goal of decisionmakingis not an internally held judgment about what would be a pos-sible course of action, or even a choice for a specific action; it isthe process starting with option generation in response to acertain situation, aiming toward the performance of a specificaction: the action is therefore considered to be part of thedecision-making process.

There are several advantages of this model. First of all, it provides astructural account of decision making that can be used to categorizeand characterize different processes in terms of their formal role. Takea specific motor representation, such as ‘eating an ice-cream’. In a pro-cess of decision making, this representation can be either a chosengoal, a mere option, a likely plan, or an actual outcome— all dependingon the role the representation plays in the process. The stage modeloffers a useful way of distinguishing such different roles or aspects ofthe process as a whole. Second, using a stage model offers ways forforensic psychiatry to connect to a strong body of existing empirical –to a large extent neuropsychological – knowledge on decision makingand its dysfunctions (Meynen, 2013). Connecting this empirical workto what we know about different mental disorders could contribute tothe development of more fine-grained assessments of criminal respon-sibility: herein lies the main relevance of our account for the field of

Please cite this article as: Kalis, A., &Meynen, G.,Mental disorder and legal rJournal of Law and Psychiatry (2014), http://dx.doi.org/10.1016/j.ijlp.2014

forensic psychiatry. Third, the model emphasizes the significance ofthe generation of options. In evaluating decisions we may too easilyfocus on picking an option, thereby overlooking that the basis for eachdecision is in the option generation. This is particularly important,since many mental disorders may affect this stage (see next section).Fourth, the model provides a structured way of assessing a decision-making process in an individual, for instance, during a psychiatric eval-uation of a defendant. Fifth, themodel shows that problems occurring inan earlier stagemay affect later stages, in which theymay ormay not becorrected. This is an advantage of ourmodel, because it enables a psychi-atrist to clearly evaluate how a particular course of action started. Final-ly, themodel suggests a certain hierarchywith respect to disturbances ofdecision making: the earlier the disturbances occur, we suggest, themore plausible it is to excuse a person who violated a moral or legalobligation. This means that in forensic practice, the clearest examplesof successful insanity defenses are derived from disturbances in thefirst phase. This should become clear in Sections 3 and 4.

3. Dysfunction in different stages

In this section, we elaborate further on our proposal by discussing:what dysfunction would look like in each of the three stages, howsuch dysfunctions might affect the decision-making process as awhole, and to what extent dysfunctions in this stage are still amenableto rational or conscious6 control. The latter point is relevant becausethe main question in forensic psychiatric evaluations is how likely it isthat a psychiatric disorder undermined the defendant's control overthe unlawful act. In order to determine this, it is important to assess towhat extent defendants could (rationally or consciously) intervene inthe outcome of the different stages of their decision-making process,and thus in their action. In this respect, it must be noted that from acriminal law point of view, it is the action that is ultimately at stake,not the options generated or even the decision made. Actions can becriminal, while perceiving something as an option, ormaking a decision,is itself never criminal in nature. Therefore, the interesting question forour aim is how distortions in different phases of the decision-makingprocess can lead to the performance of a criminal act.

3.1. Stage 1: Option generation

Various mental disorders decrease a patient's capability to see thoseoptions for action thatmost people can easily see, or they lead one to seeoptions for action that most people would not see as options. For in-stance, peoplewho suffer fromhallucinationsmay perceive phenomenathat are not actually there. They may see a door where others see

esponsibility: The relevance of stages of decisionmaking, International.02.034

4 A. Kalis, G. Meynen / International Journal of Law and Psychiatry xxx (2014) xxx–xxx

nothingmore than a brickwall: psychotic personsmay perceive optionsthat others do not. At the same time, other options may not be generat-ed: a paranoid patient may not generate the option to go to the police,because he has the delusional conviction that the police are againsthim. Such distortions in option generationmight be related to attentiondeficits and deficits in the processing of external stimuli often associatedwith schizophrenia (Braff, 1993).

People suffering from kleptomania or pyromania may generate op-tions that are not the result of delusions or hallucinations, but are stillnot the kind of options that people usually would consider; due to cer-tain obsessions, people may continuously generate behavioral optionsabout stealing, or setting fire to a house (still it looks like these may beamenable to a certain amount of conscious control, see next phase).Other problems with option generation may also occur. Some psychiat-ric patients see so few or so many options for action that it obstructstheir general ability to act in a goal-directed manner. In depression, forexample, patients may see relatively few options for action in thinkingabout how to solve everyday problems (D'Zurilla, Chang, Nottingham,& Faccini, 1998). In dementia or in some forms of apathy, option gener-ationmay almost completely come to a halt. On the other hand, patientsin amanic state often seem to generate optionswithout a pause; this canmake it very hard to give options the attention they deserve, to evaluatethem thoroughly, and therefore, to attain some level of behavioral focus.

Another problem at the level of option generation is irrational eval-uation of options: immediately seeing options asmore or less attractivethan is rationally justifiable. Although explicit evaluation of optionsbelongs to the option selection stage, it is important to note that theproblem of irrational evaluation might already start in the stage ofoption generation. Options occur to a decision maker not as neutralpossibilities, but as possibilities with a certain (affective) value (Kaliset al., 2008). For example, in depression, none of one's options maypresent themselves as interesting or appealing (Meynen, 2011): in eval-uating options depressed patients may focus more strongly on negativecharacteristics due to attentional biases (Gotlib, Krasnoperova, Yue, &Joormann, 2004;Mogg, Bradley, &Williams, 1995). Alternatively, some-one in a manic state might attach unreasonable high value to options(for an overview on decision-making dysfunctions in mania, seeMurphy et al., 2001). Patientsmay feel that a certain option is extremelypromising or brilliant: an option may ‘strike’ a person as particularlyinteresting. This value can, at least in principle, be modified in laterstages by reflection and/or changing circumstances. Of course, whethersomeone eventually translates such ‘irrational values attached tooptions’ into action depends on the second and third phase of thedecision-making process. But the decision-making process mightalready be skewed as soon as options are being generated, making itmore difficult to adjust this in later stages.

An important question in this context is: to what extent do peoplehave control over the options they generate? Note that the process ofgenerating options is usually described in terms that are relatively pas-sive: options present themselves to us, they occur to us, we get a certainidea, we see possibilities. This suggests that even in cases of perfectlyhealthy decision making, we do not experience option generation as aprocess that is amenable to conscious intervention. Of course, we canactively engage in thinking about possible things we could do: we cando more than just sit and wait for options to pop into our minds. Buteven though we can actively try to come up with options for actions,we do not know which options we will generate until we have alreadygenerated them. This is a conceptual, not an empirical, point about op-tion generation: option generation differs fromother aspects of decisionmaking in that we do not know which options we are generating andwhich options we are not generating or seeing, when we are engagedin the very process of generating them. Alternatively, in selecting op-tions, or initiating decisions, we do have knowledge of the materialwe are working with (we do know which options we are decidingbetween). This difference is important for studying dysfunctions of op-tion generation for the following reason: it might show that when the

Please cite this article as: Kalis, A., &Meynen, G.,Mental disorder and legal rJournal of Law and Psychiatry (2014), http://dx.doi.org/10.1016/j.ijlp.2014

process of option generation is disturbed somehow, there are onlyfew – if any – strategies available to a decision maker for correctingsuch disturbances on their own. To put it crudely: we do not knowwhich options we do not see. Also, if some dangerous option presentsitself as highly attractive, (almost) everything depends on stage 2 and3 for adjusting the process in order to avoid a disaster. If the input inboth of these stages is not sufficient enough to correct the option as itpresented itself – as may occur in delusional states – the dangerous op-tion might materialize in concrete action (see Fig. 2).

Indirect control, however, might still be possible, at least to someextent: a doctor or therapist might tell depressed patients, repeatedlyand convincingly, that they really could take a music class, and thismight give them an idea that they would not have thought of by them-selves. Others might also convince a manic person that certain optionsare not really as feasible or brilliant as the person thinks they are.

A person could thus receive external help for correcting certain dys-functions in option generation. But without such help, requiring some-one to come up with a certain option on their own accord is likerequiring someone to read a book with their eyes closed. As we willshow, the limited possibilities to consciously intervene in your ownspontaneous option-generation process are what distinguish dysfunc-tions in this stage from dysfunctions at other stages of the decision-making process: a difference that could be relevant for assessments ofcriminal responsibility.

3.2. Stage 2: Option selection

Dysfunctional decision making in the stage of option selection canmanifest itself as indecision, or as irrational choice (not choosing the op-tion that is evaluated as most suitable, given one's goals). Many disor-ders may interfere with the process of option selection. First of all,certain cognitive capacities may be lacking, as may be the case in de-mentia. In cases of dementia, people lose the capacity to contemplateoptions in a rational way; peoplemay even forget certain options duringthe deliberation in the selection phase. Reflection on the various optionsrequires that they remain present before the mind's eye. Other peoplemay have attention problems (as in ADHD): they might focus on onespecific option without paying sufficient attention to its consequences,and to the possibly better consequences of other courses of action. Alter-natively, people may have difficulty with selecting one of the options.People suffering from OCD or depression may be indecisive for a longperiod time, before being able to select an option. People sufferingfrom hypomania may not be able to evaluate options properly in theselection phase: their thinking may be associative rather than rational;in addition, they may be unreasonably positive about what conse-quences certain actionsmay have, and therefore select risky, dangerousor criminal options (see Fig. 3).

To what extent are dysfunctions in the phase of option selectionamenable to control, or conscious intervention? Some problems in op-tion selection might be more amenable to conscious intervention thanproblems in option generation. For example, even if one is heavilytempted to decide to have a drink, one might at the same time beaware that choosing this option would be a very unwise decision.Such awareness can, in some situations, be put to work in order todecide on a different option instead. One might do this either byattempting to use pure willpower (direct control), or by employingtricks for manipulating one's own behavior (indirect control, see Mele,1987). Consider alcoholism: the fact that alcohol addicts have allegedlysome control over selecting their options (even though the urge to drinkmight be very strong), may make people reluctant to consider alcohol-ism grounds for exculpation (Fingarette, 1989). It may be argued thatone still ultimately decides whether or not to ‘give in’ to the urge todrink. Furthermore it could be argued that by modifying their situation,persons can ensure that theywill (not)make a certain choice. By stayingat home, alcohol addicts can ensure that they will not decide to entertheir favorite bar in town on the spur of the moment. In therapy, and

esponsibility: The relevance of stages of decisionmaking, International.02.034

Fig. 2. Dysfunctional option generation leading to an unlawful act.

5A. Kalis, G. Meynen / International Journal of Law and Psychiatry xxx (2014) xxx–xxx

with the help of others, one can decide not towash one's hands at a cer-tain moment: the obsessive thought still arises, and the usual responseof the patient (urge towash one's hands) still arises, but patients canputeffort in not selecting this option as their course of action. Training thiscapacity to intervene in one's own option-selection process, stands atthe base of many forms of cognitive behavioral treatment of mental dis-orders (Pickard, 2011). Of course patients will not always succeed: evenin healthy decision making, the capacity to intervene is certainly not al-mighty; in cases of mental disorder, it might be precisely this capacitythat is undermined. Also, this kind of control is not always possible: incases of mania or dementia, for example, persons might not even beaware that they are making an irrational decision.

3.3. Stage 3: Action initiation

Problems at the level of action initiation can manifest themselves asnot doing thatwhat you chose to do, or doing something too early or late,or doing something different from what you chose to do. To give an ex-ample of dysfunction in this stage: the syndrome of Gilles de la Touretteconsists of tics. These tics are, at least in part, actions that are initiatedeven before a decision-making process has been started, let alone prop-erly completed (Lang, 1991; Singer, 2011; Verdellen et al., 2008). Otherimpulse control disordersmay also be considered not only as exhibitingproblems with the selection phase, but also as showing too early actioninitiation. People act before they have properly selected what a goodcourse of action would be. Problems of action initiation can also be amanifestation of certain kinds of apathy in the sense of not translatingchoices into action. For instance in negative symptoms in schizophrenia,people might make choices but nevertheless remain inert (Morris,Rushe,Woodruffe, &Murray, 1995). Addiction can also lead to problemsof action initiation: an alcohol addict can repeatedly decide to stopdrinking, but nevertheless fail to actually translate this decision into

Fig. 3. Dysfunctional option selecti

Please cite this article as: Kalis, A., &Meynen, G.,Mental disorder and legal rJournal of Law and Psychiatry (2014), http://dx.doi.org/10.1016/j.ijlp.2014

real action, ‘I will stop tomorrow’ (whether addictive behavior is the re-sult of dysfunctional option selection or action initiation, thus depends onthe question whether addicts makes a decision to drink, or whether theyhave problems acting on their decision not to drink). Dysfunctions of ac-tion initiation can lead to criminal behavior as well. Taking illegal drugs,despite the decision not to do it anymore, could be an example (Fig. 4).

Concerning the role of conscious control, dysfunctions of actioninitiation seem comparable to dysfunctions of option selection.Transforming a decision into actual behavior is brought about by pro-cesses that are to some extent amenable to conscious intervention(Bratman, 1987; Bratman, 1997). Of course, this does not mean thatthese processes are always or necessarily characterized by control: dys-functions can manifest themselves precisely as a loss of such control, asin impulse control disorders (Sebanz & Prinz, 2006). However, as withoption selection, a person with such problems can make use of certaintricks or strategies that might help him overcome or correct such dys-functions. Even for a disorder such as Tourette's syndrome, which is atleast in part characterized by involuntariness (Singer, 2011) thatseems to preclude the possibility of control, cognitive behavioral thera-py has been developed in which people learn to better resist the occur-rence of a tic. The exact process of tic-formation and tic-control is stillunclear, but it looks like gaining control over the tics is, at least tosome extent, possible (Verdellen et al., 2008). The issue of control andresponsibility is also hotly debated with regard to addiction (Baler &Volkow, 2006; Kalis, 2011; Mele, 2002). In at least some cases, personswith an addictionmight be able tomanipulate their situation in order toensure that they will stick to their decision, and not have another drink.Willpower could also play a role here, although research on addictionshows that this is usually precisely the process that is affected in casesof addiction (Bechara, 2005).

It is clear from the examples derived from psychopathologydiscussed above that symptoms or disorders may affect different stages

on leading to an unlawful act.

esponsibility: The relevance of stages of decisionmaking, International.02.034

Fig. 4. Dysfunctional action initiation leading to an unlawful act (the lawful option has been selected, but is not translated into action, while an unlawful act is – immediately – translatedinto action).

6 A. Kalis, G. Meynen / International Journal of Law and Psychiatry xxx (2014) xxx–xxx

of the decision-making process simultaneously. For example, apathymay affect both option generation and action initiation. And a manicepisodemay result in increased option generation, speedy option selec-tion, and immediate initiation of the action. For instance, the idea comesup to invest one's money in a certain company (the option appears asbrilliant), the option is immediately selected, and translated into action.Because of such sequences, patients experiencing manic episodes maylose a lot (such as their money, their friends, and their work), in avery short period of time.

4. Back to the insanity defense

So far, we have seen that the stage of option generation seems lessamenable to control and self-correction than the other stages of thedecision-making process. The basic point is that it is very hard makeyourself see options you do not see, at least not without external help.On the other hand, you can make yourself decide or do somethingthat you do not consider to be that most attractive option: consciousintervention is at least sometimes possible in later stages of decisionmaking, and it amounts to selecting or initiating an action based onthe different options thatwere generated in the first stage. This strategyof ‘going back’ to the first stage to correct one's own decision making isnot available, if it precisely the first stage that is dysfunctional. Thismakes option generation especially relevant for the questionwhichdys-functions are important in the context of an insanity defense. Whatever‘goes wrong’ in the process of generating options may affect otherstages, such as option selection, action initiation, and action evaluation.

The framework presented so far is at least compatible with the factthat it is often psychotic disorders that lead to a successful insanity de-fense (Nestor & Haycock, 1997), rather than disorders like depression,addiction, or OCD. Psychosis could result in the generation of ‘insane’options that could lead to ‘insane’ actions. If there are profound prob-lems in generating options, then the patient can often do very little asfar as prevention of criminal behavior is concerned. In fact, the disordersthat are generally thought to be relevant for the question of criminalresponsibility are those disorders in which it is most difficult for theperson to distance oneself from one's dysfunctions (like delusions andhallucinations). As we have tried to show above, although we realizethat we have not provided a full explanation, this might be becausethese disorders specifically affect processes of option generation,which are, even in healthy decisionmakers, much less amenable to con-trol than other processes of decision making.

Consider, for instance, the case of Daniel M'Naghten. His legal case,eventually, resulted in the development of the M'Naghten rule, thelegal standard most widely used in Anglo-American jurisdictions (seeIntroduction). This rulemight strike us asmainly related to the selectionphase, in which the nature of the options receives our full attention, andwe may reject certain options because they would be ‘wrong’. But the

Please cite this article as: Kalis, A., &Meynen, G.,Mental disorder and legal rJournal of Law and Psychiatry (2014), http://dx.doi.org/10.1016/j.ijlp.2014

case itself suggests otherwise. Daniel M'Naghten was deluded. He wasconvinced that the Tories (the political party) were persecuting him,and, therefore, he tried to assassinate the Prime Minister, Sir RobertPeel (he apparently mistook the secretary to the Prime Minister(Edward Drummond) for the Prime Minister, and killed him). Now,what actually seemed to have happened here is, first of all, that due toa deluded and distorted perception of reality, M'Naghten started tosee options of a certain kind: perhaps killing the Prime Minister, per-haps killing other people — we do not know (for sure). In any case, heeventually came up with a plan. These options arose spontaneously,from a profoundly distorted perception of reality. Now, that samedistortion probably also had its implications for the second stage ofdecision making: the selection of options was evaluated against thebackground of his deluded view about reality. And he selected, it isassumed, one of the options, namely to kill the Prime Minister, SirRobert Peel, in a certain way at a certain moment in time (Robinson,1998). In this selection stage, there does not seem to be room for con-templations like: Is it really true that I am being persecuted by theTories? The nature of delusions is such that beliefs produced by themare in fact unshakable, and all the alternative options are likely to arisewithin the constraints of the person's distorted perception. As far aswe know, M'Naghten did not suffer from impulse problems. Therefore,it might be that the third stage was not affected by psychopathology.But we can also doubt, of course, whether the threshold of committingsuch an act was not also lowered by his mental disorder. To conclude,M'Naghten's psychotic state probably primarily manifested itself inthe fact that he generated a 'crazy' range of options for action.

Let us consider another psychotic phenomenon that can be groundfor exculpation: commanding voices (Donohue, Arya, Fitch, & Hammen,2008). Commanding voices occur in an estimated 33–74% of peoplewho experience hearing voices (Bucci et al., 2013), and they may pro-foundly influence a person's behavior (Meynen, 2013; Reynolds &Scragg, 2010; Shawyer et al., 2008). From a criminal law perspective,two issues regarding such phenomena are important: the contents ofthe command and the possibility on the part of the patient to disobeythe command. In some (rare) cases, it is impossible for a patient notto obey a command given by a voice. As long as the contents of the com-mands are harmless, like, ‘Make some coffee!’ or ‘Go outside!’, they areunlikely to affect a person's behavior in a legally relevant way, even ifthe patient is not able to disobey the command. But if the contents areharmful, such as ‘Kill your neighbor!’, and if the patient cannot helpbut obey the command, these voices may become highly relevantfrom a criminal law perspective. Consider a patient suffering fromschizophrenia who acts in accordance with such a voice that cannotbe disobeyed. As a result, the neighbor is attacked and severely wound-ed. Let's consider the patient's decision making in view of the threestages. The option was generated by the psychopathological phenome-non, namely the voice (stage 1). In addition, the nature of this

esponsibility: The relevance of stages of decisionmaking, International.02.034

7A. Kalis, G. Meynen / International Journal of Law and Psychiatry xxx (2014) xxx–xxx

commanding voicewas such that the patient did not have the decisionalspace to select another option or to block this particular, harmful option(stage 2). Finally, since no problems exist with action initiation in thispatient,7 the selected option is translated into harmful action. The vari-ous aspects of (compliancewith) command hallucinations need furtherclarification (Bucci et al., 2013; Shawyer et al., 2008). Still, from thestaged decision-making perspective we have outlined, the significantand intriguing feature of these phenomena is that the generation ofthe option implies the selection of that option. Unlike in normal decisionmaking, there is, apparently, no room for not selecting this option: it hasto be selected. This is due to theway the option presents itself to the per-son, namely as a command that cannot be disobeyed. In conclusion, thispsychopathological phenomenon appears to affect (at least) two stagessimultaneously; such a phenomenon is (potentially) dangerous, due tothe fact that its contents can be translated into action immediately,without any control on the part of the patient.

5. Conclusions

In this paper, we have tried to show how a stage model of decisionmaking could provide a useful theoretical framework for specifyingwhy we excuse people who suffer from a mental disorder – at leastin certain circumstances – for their actions. We have argued that dif-ferent psychiatric disorders can manifest themselves as dysfunctionsin different stages in the decision-making process, and that somestages of this process are less amenable to control – conscious inter-vention – than others. In particular, the stage of option generation(stage 1) was shown to be difficult to control directly: althoughsome forms of indirect control are possible with external help, it isnot up to us whether we see something as an option for action ornot. This also means that if this stage of the decision-making processoperates in a dysfunctional way, it is hard to correct such dysfunctionsby direct exercise of control or willpower: even in healthy decisionmaking, willpower has no direct role in option generation. Willpowercan only be effective in preventing harmful actions by stopping anoption from being translated into action (stages 2 and 3). However,stopping an action from being translated into actionmight be difficult,if the disorder has also affected these stages. This could also explainwhy insanity defenses are mostly used in cases of psychosis: insofaras these disorders affect decision making, they seem to affect boththe option generation stage and the option selection stage. We there-fore suggest that assessing option-generation dysfunctions is highlyrelevant for judgments of moral and criminal responsibility. Also,given the aforementioned difficulties in directly controlling optiongeneration, further research is needed to study possibilities for exter-nal guidance of option-generation processes in cases of dysfunction,such as offering suggestions for action or adjusting the prima facieevaluations of options in the context of therapy.

The proposed model of stages in decision making may help to evalu-ate the impact of amental disorder on a patient's or defendant's decision-making process. In this way, we suggest, the model could inform moraland legal judgments about a defendant's criminal responsibility.

Acknowledgments

The research of Kalis was supported by a grant from the VolkswagenFoundation for the project “Irrationality as dysfunctional option gener-ation” within the European Platform for Life Sciences, Mind Sciences,and the Humanities.

7 Although, some researchers suggest that impulsivity may be a factor in compliancewith such voices (Bucci et al., 2013).

Please cite this article as: Kalis, A., &Meynen, G.,Mental disorder and legal rJournal of Law and Psychiatry (2014), http://dx.doi.org/10.1016/j.ijlp.2014

References

American Law Institute (1962). Model penal code. Philadelphia, PA: American LawInstitute.

Baler, R. D., & Volkow, N. D. (2006). Drug addiction: The neurobiology of disrupted self-control. Trends in Molecular Medicine, 12(12), 559–566.

Bechara, A. (2005). Decision making, impulse control and loss of willpower to resistdrugs: A neurocognitive perspective. Nature Neuroscience, 8(11), 1458–1463.

Borum, R., & Fulero, S. M. (1999). Empirical research on the insanity defense andattempted reforms: Evidence toward informed policy. Law and Human Behavior,23(3), 375–394.

Braff, D. L. (1993). Information processing and attention dysfunctions in schizophrenia.Schizophrenia Bulletin, 19(2), 233–259.

Bratman, M. (1987). Intentions, plans, and practical reason. Cambridge, MA: HarvardUniversity Press.

Bratman, M. (1997). Responsibility and planning. Journal of Ethics, 1(1), 27–43.Bucci, S., Birchwood, M., Twist, L., Tarrier, N., Emsley, R., & Haddock, G. (2013). Predicting

compliance with command hallucinations: Anger, impulsivity and appraisals ofvoices' power and intent. Schizophrenia Research, 147(1), 163–168.

Buchanan, A. (2000). Psychiatric aspects of justification, excuse and mitigation. London, UK:Jessica Kingsley.

D'Zurilla, T. J., Chang, E. C., Nottingham, E. J., & Faccini, L. (1998). Social problem-solvingdeficits and hopelessness, depression, and suicidal risk in college students andpsychiatric inpatients. Journal of Clinical Psychology, 54(8), 1091–1107.

Donohue, A., Arya, V., Fitch, L., & Hammen, D. (2008). Legal insanity: Assessment of theinability to refrain. Psychiatry (Edgmont), 5(3), 58–66.

Elliott, C. (1996). The rules of insanity. Moral responsibility and the mentally ill offender.Albany, NY: SUNY Press.

Ernst, M., & Paulus, M. P. (2005). Neurobiology of decision-making: A selective reviewfrom a neurocognitive and clinical perspective. Biological Psychiatry, 58, 597–604.

Fingarette, H. (1989). Heavy drinking: The myth of alcoholism as a disease. Berkeley, CA:University of California Press.

Gotlib, I. H., Krasnoperova, E., Yue, D. N., & Joormann, J. (2004). Attentional biases fornegative interpersonal stimuli in clinical depression. Journal of Abnormal Psychology,113(1), 127–135.

Heckhausen, H., & Gollwitzer, P. M. (1987). Thought contents and cognitive functioning inmotivational versus volitional states of mind. Motivation & Emotion, 11, 101–120.

Heckhausen, J., & Heckhausen, H. (Eds.). (2008). Motivation and action. Cambridge, MA:Cambridge University Press.

Kalis, A. (2011). Failures of agency: Irrational behavior and self-understanding. Lanham,MD: Lexington Books.

Kalis, A., Mojzisch, A., Schweizer, T. S., & Kaiser, S. (2008). Weakness of will, akrasia, andthe neuropsychiatry of decision making: An interdisciplinary perspective. Cognitive,Affective, & Behavioral Neuroscience, 8(4), 402–417.

Kalis, A., Kaiser, S., & Mojzisch, A. (2013). Why we should talk about option generation indecision-making research. Frontiers in Psychology, 4, 555.

Lang, A. (1991). Patient perception of tics and other movement disorders. Neurology,41(2), 223–228.

Mele, A. R. (1987). Irrationality: An essay on akrasia, self-deception, and self-control. NewYork, NY: Oxford University Press.

Mele, A. R. (2002). Akratics and addicts. American Philosophical Quarterly, 39(2), 153–167.Meynen, G. (2010). Free will and mental disorder: Exploring the relationship. Theoretical

Medicine and Bioethics, 31(6), 429–443.Meynen, G. (2011). Depression, possibilities, and competence: A phenomenological

perspective. Theoretical Medicine and Bioethics, 32(3), 181–193.Meynen, G. (2012). An ethical framework for assessments of criminal responsibility:

Applying Susan Wolf's account of sanity to forensic psychiatry. International Journalof Law and Psychiatry, 35(4), 298–304.

Meynen, G. (2013). A neurolaw perspective on psychiatric assessments of criminalresponsibility: Decision-making, mental disorder, and the brain. InternationalJournal of Law and Psychiatry, 36, 93–99.

Mogg, K., Bradley, B. P., & Williams, R. (1995). Attentional bias in anxiety and depression:The role of awareness. British Journal of Clinical Psychology, 34(1), 17–36.

Morris, R. G., Rushe, T., Woodruffe, P. W. R., & Murray, R. M. (1995). Problem solving inschizophrenia: A specific deficit in planning ability. Schizophrenia Research, 14(3),235–246.

Morse, S. J. (2003). Diminished rationality, diminished responsibility. Ohio State Journal ofCriminal Law, 1, 289–308.

Morse, S. J. (2011). Mental disorder and criminal law. The Journal of Criminal Law &Criminology, 101(3), 885–968.

Murphy, F. C., Rubinsztein, J. S., Michael, A., Rogers, R. D., Robbins, T. W., Paykel, E. S., et al.(2001). Decision-making cognition in mania and depression. Psychological Medicine,31(4), 679–693.

Nestor, P. G., & Haycock, J. (1997). Not guilty by reason of insanity of murder: Clinical andneuropsychological characteristics. Journal of the American Academy of Psychiatry andthe Law Online, 25(2), 161–171.

Pickard, H. (2011). Responsibility without blame: Empathy and the effective treatment ofpersonality disorder. Philosophy, Psychiatry, Psychology, 18(3), 209–222.

Reynolds, N., & Scragg, P. (2010). Compliance with command hallucinations: The role ofpower in relation to the voice, and social rank in relation to the voice and others.The Journal of Forensic Psychiatry & Psychology, 21(1), 121–138.

Robinson, D. N. (1998).Wild beasts and idle humours: The insanity defense from antiquity tothe present. Cambridge, MA: Harvard University Press.

Rogers, R., & Shuman, D. W. (2005). Fundamentals of forensic practice: Mental health andcriminal law. New York, NY: Springer.

Sebanz, N., & Prinz, W. (Eds.). (2006). Disorders of volition. Cambridge, MA: MIT Press.

esponsibility: The relevance of stages of decisionmaking, International.02.034

8 A. Kalis, G. Meynen / International Journal of Law and Psychiatry xxx (2014) xxx–xxx

Shawyer, F., Mackinnon, A., Farhall, J., Sims, E., Blaney, S., Yardley, P., et al. (2008). Actingon harmful command hallucinations in psychotic disorders: An integrative approach.Journal of Nervous and Mental Disease, 196(5), 390–398.

Simon, R. J., & Ahn-Redding, H. (2006). The insanity defense, the world over. Lanham, MD:Lexington Books.

Singer, H. S. (2011). Tourette syndrome and other tic disorders. Handbook of ClinicalNeurology, 100, 641–657.

Smaldino, P., & Richerson, P. (2012). The origins of options. Frontiers in Neuroscience,6(50), 1–11.

Please cite this article as: Kalis, A., &Meynen, G.,Mental disorder and legal rJournal of Law and Psychiatry (2014), http://dx.doi.org/10.1016/j.ijlp.2014

Verdellen, C. W., Hoogduin, C. A., Kato, B. S., Keijsers, G. P., Cath, D. C., & Hoijtink, H. B.(2008). Habituation of premonitory sensations during exposure and responseprevention treatment in Tourette's syndrome. Behavior Modification, 32(2),215–227.

Vincent, N. A. (2011). A structured taxonomy of responsibility concepts. In N. A. Vincent, I.van de Poel, & J. van den Hoven (Eds.), Moral Responsibility: beyond free will & deter-minism (pp. 15–35). Dordrecht: Springer.

Wallace, R. J. (1994). Responsibility and the moral sentiments. Cambridge, MA: HarvardUniversity Press.

esponsibility: The relevance of stages of decisionmaking, International.02.034