Thinking differently ...

... about cognitive distortions

Robert McGrath and colleagues’ review of the core treatment targets of sexual offender treatment programmes (McGrath, Cumming, & Burchard, 2003) suggests that most treatment programmes

  • fail to target criminogenic risk factors and, instead,
  • direct too many resources toward characteristics that have little or no demonstrated empirical relationship with recidivism (e.g. victim awareness and empathy).

We have argued that, if responsibility is understood as making internal, stable and intentional attributions for one’s crimes (‘I did it because I wanted to), then taking full responsibility for one’s crimes is also overrated as a treatment goal – and may even be counterproductive in some instances.

This does not mean that we advocate that offenders be taught or encouraged to make excuses in treatment (as this seems to be a natural process for most anyway), only that those working with offenders rethink their assumptions about excuse-making.

[... ... ...]

It is important to note that, unlike with recidivism, there is a clear, empirical link
between minimization of responsibility and treatment attrition rates [...].

This finding is sometimes interpreted as a rationale for denying treatment to offenders who will not take responsibility for their crimes. As getting offenders to accept responsibility better for their crimes is typically considered a treatment goal, however, this would be

  • ‘tantamount to requiring them to : : : cure themselves before they can receive treatment’ (Schneider & Wright, 2004, p. 7).

Instead, the finding that minimization is linked to programme attrition could more usefully be reinterpreted as a rationale for changing the accepted styles of dealing with minimization in treatment

  • (see especially Beyko & Wong, 2005; Maletzky, 1996).

For instance, it is already generally accepted in the field of offender rehabilitation
that confrontational methods are not helpful in assisting motivation and change
(Murphy & Baxter, 1997). The series of studies conducted by Marshall and associates
[... ...] indicate that a warm, genuine, motivational style is more effective at promoting change than a confrontational, judgmental style [...].


Most notable in this regard is the ‘good lives’ model developed by Tony Ward and his colleagues (see e.g. Ward & Brown, 2004) that expands the rehabilitative focus from targeting deficits to building strengths.

In this final section, we set out some additional alternatives to working with issues of personal responsibility in research and practice.

Consider different ways of taking responsibility

A probable reaction to what we have written is the question: ‘How can people
change if they do not accept responsibility for the acts they have committed?’

One response is that taking responsibility for one’s actions is about more than attributions for past behavior.

Bovens (1998) differentiates between passive and active responsibility.

  • Whereas passive responsibility means holding someone responsible for something they have done in the past,
  • active responsibility means the virtue of taking responsibility for putting things right for the future (...).
    Active responsibility is future oriented and forward thinking, focusing on what needs to be done in order to make good or make amends or make it right (...). 

Research on ex-prisoners who have been able successfully to desist from crime finds these active responsibility themes to be far more prevalent than internal, stable and intentional attributions for one’s past crimes (Maruna, 2001; Stefanakis, 1998).

Brickman and colleagues (1982) suggested a third way in regard to the vexing
internalizing–externalizing debate. In what they refer to as a compensatory model of
, individuals do not blame themselves for their problems, but hold
themselves responsible for the solution to the problems.

Brickman et al. (1982, p. 372) quoted the Reverend Jesse Jackson’s various slogans as being representative of this model of responsibility

  • (e.g. ‘You are not responsible for being down, but you are responsible for getting up’).

Maruna’s (2001) desistance research suggested that this compensatory model characterizes the self-narratives of successfully desisting exconvicts.
He concluded that, although it may be therapeutic for a person to locate the
roots of one’s problems in the social environment (disadvantage, inequality,
victimization), to desist one might need to internalize responsibility for overcoming
these obstacles (see also Braithwaite & Braithwaite, 2001).

Dividing the concept of responsibility into ‘blame for the past’ and ‘control over the future’ then, might be a valuable therapeutic tool (...]. 

Consider account dimensions besides internality

As explanatory style research has progressed in other areas, the internality–externality
dimension central to the locus of control concept has become of less interest to

  • ‘It has more inconsistent correlates than do stability or globality,
  • it is less reliably assessed and
  • there are theoretical grounds for doubting that it has a direct impact on expectations per se’ (Peterson, 2000, p.48).

In particular, future neutralization research needs to transcend the overly simplistic and long passé (...) notion that an internal locus of control is always better than an
external locus of control, and develop a more complex understanding of neutralizations.

One suggestion from the psychological literature is to focus more on other dimensions
of attributions, such as

  • stability,
  • globality,
  • intentionality and
  • controllability.

These other dimensions have been found to be better predictors of a variety of other outcomes (...).

For instance, excuses and justifications that rely upon

  • hightly stable and global attributions (e.g. ‘That is just the way the world works;’ ‘This is just who I am’) and
  • attributions of a hostile nature (e.g. ‘It is because everyone is against me’)

may be most likely to be associated with persistent criminality (see Maruna, 2004).

Excuses that separate past offending behaviour from the individual’s core self (e.g. ‘It was a complete accident’) may be more commonly associated with maintaining desistance from crime.

These remain important, lingering questions, not just for their value in working with
ex-offenders, but also because they allow us better to understand desistance or
persistence from the individual offender’s point of view.

Treat excuses as the identification of risk factors

Process evaluations of the dynamics of therapeutic encounters (e.g. Marshall,
Fernandez, et al., 2003) suggest that counsellors should listen with interest to the
messy, realistic, explanations that clients offer, rather than rejecting these automatically
as cover-ups.

Honouring accounts in this way not only builds trust and promotes cooperation, it also generates more valuable material for therapeutic work. That is, statements of cause that may typically have been categorized as excuses may actually point helpfully towards criminogenic needs and dynamic risk factors.

For instance,

  • the offender who explains that ‘I did it because I was stressed’ is noting that he has poor coping skills for managing life stress.
  • The offender who thinks his offending may have been caused by loneliness is revealing that he lacks intimacy in his adult relationships and that this is a source of distress to him.

When untreated sexual offenders are asked to explain their offending, they typically
provide reasons which reflect known dynamic risk factors for this type of crime, such as

  • intimacy seeking,
  • poor emotional management or
  • impulsivity (Mann & Hollin, 2006).

Confronting and eradicating these explanations leaves both therapist and client with
little raw material for dealing with the crucial psychosocial issues surrounding the

Consider the excerpt from Salter (1988), cited above. Many of the issues she
described as externalizations, such as

  • lack of emotional intimacy,
  • loneliness and
  • emotional congruence with a child,

have now been empirically established as factors that raise the risk of recidivism (Hanson & Morton-Bourgon, 2005).

To eliminate such factors from an offender’s account of his crimes could dangerously affect his ability to understand, recognize and manage acute changes in risk in the future.

After all, the self-statement ‘I did it because I wanted to’ meets the standards of being an internal, stable, attribution of cause, but is futile as a basis for awareness of how to avoid
offending in the future (see Mann & Hollin, 2006).

Probe cognitive style beyond offence accounts

Schneider and Wright (2004, p. 15) argued that

  • ‘The primary vehicle for assessing and modifying offenders’ cognitions is likely to be found in the explanations provided by offenders to account for their offences’.

Very much on the contrary, Hanson and Morton-Bourgon (2005, p. 1159) concluded from their recent meta-analysis that it is possible that evaluators and treatment providers looking for risk factors

  • ‘have little to gain from listening to offenders’ attempts to justify their transgressions’.

Of these two assessments, we are persuaded that the latter is probably the safer conclusion.

Of all the cognitions available to treatment professionals, offence accounts appear to be the most ‘loaded’ – for legal reasons and purposes of self-definition and identity
management. Why prioritize these accounts over any other in assessing client locus
of control or self-efficacy?

In the last 15 years, there have been three major reviews of cognitive factors in
sexual offending

  • (Drieschner & Lange, 1999; Segal & Stermac, 1990; Ward et al., 1997).

All three reviews concluded that studies of sexual offenders’ cognition have focused far
too much on measurement of excuses, justifications and offence-supportive attitudes, at
the expense of differentiating cognitive structures, processes and outputs. 

[... ... ...]

Ward suggests that instead of focusing on the symptoms of distorted thinking, correctional counsellors should look to the source or the cognitive schemata underlying these patterns of belief.

Likewise, Beech and Mann (2002, p. 268) describe a schema-based treatment programme that focuses not just on offence-justifying attitudes, but also on underlying self-understandings, motivations and implicit beliefs.

[... ...]

A shift in focus from rationalizations to [a new] schema would require more of an attempt to tease out the broad views individuals have of the self, others and the social world and the way these are embedded in practices and ways of living [...].

Instead of aiming to eliminate neutralizations, it is suggested that rehabilitative interventions should primarily seek to assist the offender in understanding his characteristic thinking patterns where they may have contributed to his eventual selection of antisocial behaviour as his response in a given situation. By focusing solely on post-offence cognitive products such as excuses, treatment providers risk ignoring the stable underlying cognitive structures such as schemas and leave clients open to future processing errors.

Separate offence-supportive attitudes from rationalizations

An attitude is defined as an orientation towards an object or concept, that comprises

  • a cognitive component,
  • an evaluative component,
  • an affective component and
  • a disposition for action.

Offence-supportive attitudes and beliefs are defined as enduring, non-situation-specific (Blumenthal et al., 1999) attitudes which justify or support sexual offending.

Offenders’ attitudes have been extensively studied, usually under the term
cognitive distortions (Abel et al., 1989). Yet, the relationship between excuses and
attitudes is murky and has received little research attention in criminological psychology
(...) so grouping them as parallel entities is unwarranted.

Unlike excuses and rationalizations, considerable evidence exists for the relevance of such attitudes to habitual offending.

For instance, considerable research suggests that attitudes such as the belief that children enjoy sexual contact with adults, are linked to sexual offending (Hanson & Morton-Bourgon, 2005).

Furthermore, it appears that such enduring attitudes promote the likelihood that an offender will attribute provocation to their victims (Mann, et al., 2006). These beliefs have been categorized as cognitive distortions along with excuses, but they are probably a very
different phenomenon and a better target for treatment.

Differentiate between ‘good’ accounts and ‘bad’ excuses

Some work in the cognitive distortion tradition suggests that all offender accounts are
created equal: essentially any explanation is a bad explanation.

For instance, in their influential study of the thinking errors of convicted rapists, Scully and Marolla (1984) appear to categorize every causal attribution that their subjects made as either an excuse or a justification – the coding scheme being utilized apparently had no other categories.

[... ...]

This blanket treatment is unfortunate, and might be one reason why neutralization theory has failed to gain a great deal of empirical support.

It would make sense for future research to try to identify the elements that make some
accounts adaptive and others non-adaptive (see Schlenker, Pontari, & Christopher, 2001).
In particular, priority might be to identify which rationalizations are the most toxic and
separate these out from the more neutral or even benign explanations offenders hold on to in order to maintain their self-esteem.

Some contenders for the worst of the worst explanations might be

  • ‘dehumanizing one’s victims’,
  • ‘seeing the world as hostile’ or
  • ‘labelling one’s self as “naturally” deviant’.


[...] There is considerable support for the relationship between aggression and
a ‘hostile attribution bias’ (Dodge, 1993) by which aggressive school children misinterpret
social cues, perceiving external threat and aggressive antagonism under ambiguous

On the other hand, excuses such as ‘blaming one’s upbringing’ or ‘alcohol blame’ may be useful insights into a person’s risk factors and limitations.

It is an open question that would require more sensitive and sophisticated research methodologies.