Knowledge in crime and security

‘There is nothing so practical as a good theory’ – Kurt Lewin, founder of action research in social psychology.

Why do we need deliberately-designed tools for transfer and sharing of crime prevention knowledge?

Crime prevention is a highly complex activity that is challenging to do successfully. Studies of the rollout of national programmes and the replication of individual ‘success story’ projects have often revealed a significant level of ‘implementation failure’. This problem has a range of causes, including failures in project management processes. But a major issue concerns knowledge of crime prevention and how it is collected, transferred and applied.

Knowing ‘what works’ on the basis of reliable evidence from impact evaluations is vitally necessary for successful crime prevention. But to ensure delivery of good performance by practitioners, it is not enough to have this knowledge only in ‘headline’ terms (e.g. lighting reduces crime)’ – see Ekblom (2002b); Pawson (2006)).

More detailed information is needed – for example on what kind of lighting against what kind of crime. However, attempting to replicate a successful project in exact, literal detail – ‘cookbook copying’ – will also fail. This is because preventive action is very context-dependent for its success (what works in one place may not work in exactly the same form in other circumstances). It relies on practitioners intelligently following a process of identifying and solving a given crime problem, and customising generic preventive principles to activate specific causal mechanisms of prevention which fit the current context.

In fact, what we can know about crime prevention practice is much wider than ‘what works’, or ‘what is cost-effective’. Appreciating this issue requires a deeper understanding of the functions and forms of crime and security knowledge.

The functions of crime and security knowledge

Crime and security knowledge have several functions: guiding practice, facilitating delivery of programmes, informing policy, public understanding and debate, and supporting a disciplined and systematic yet imaginative approach to secure design, and crime futures.

Defining, capturing, filtering, organising, synthesising, retrieving and transferring that knowledge are all vital activities which merit investment in systematising and managing it. Research and evaluation activities both feed into the knowledge base and feed from it.

A presentation on the role of evaluation in these various functions of knowledge – especially in best practice – is here.

Forms of crime knowledge – the 9 Ks

Security-related knowledge takes several forms:

• Know-of crime: what crime is, definitions of specific offences; definitions of related concepts such as security, safety and partnership.
• Know-about crime problems: their nature, costs and wider harmful consequences for victims and society; offenders’ modus operandi, patterns and trends in criminality, risk and protective factors… and theories of causation.
• Know-what works or is theoretically/practically plausible: what methods work, against what crime problem, by what mechanisms, in what context, with what side-effects and what cost-effectiveness. More generally, when we lack ‘what works’ evidence for addressing a particular problem, the knowledge practitioners and policymakers need to fall back on what are theoretically and practically plausible intervention methods and principles.
• Know-who to involve and how: contacts for advice, potential partners and collaborators who can be mobilised as formal or informal preventers; service providers, suppliers of funds and equipment and other specific resources; and sources of wider support.
• Know-when to act: knowing the right time to make particular moves – the climate must be right, other initiatives need to be coordinated with etc.
• Know-where to target and distribute resources: in relation to the location, scale and seriousness of crime problems, and of other social action in the area.
• Know-why: covering the symbolic, emotional, ethical, cultural, political and value-laden meanings of crime and reductive action, including fairness and justice. Failure to address these issues can cause even the most rational and evidence-based actions to be rejected. The classic example is the public outrage sometimes caused by expensive sporting activities for young offenders.
• Know-how to put into practice: knowledge and skills of implementation and process models for crime prevention and security; and methodologies for research, analysis and evaluation.
• Know-knowledge and its management: its capture, assessment and evaluation, consolidation, dissemination, application and development; how to extend and improve knowledge management processes; how to make knowledge more valued in practitioner culture; learning how to learn; evolution of evolvability in cultural knowledge.

Note – this has been extended from the 7 Ks of crime knowledge.

Supporting these forms of knowledge requires frameworks to make the process systematic, rigorous and convenient.

Knowledge forms represented in the Crime Frameworks

Here are the Crime Frameworks, roughly aligned with the knowledge types.

Framework
Terms, definitions & concepts (Know-of)	Building blocks for all the Crime Frameworks – systematic, consistent, in-depth suite of terms, definitions and concepts for crime prevention and security.
5Is (Know-knowledge, Know-how)	A knowledge-management framework for capturing, assessing, synthesising, transferring and replicating good practice in crime prevention; and a process model for doing crime prevention which is a more detailed equivalent of SARA.
CLAIMED (Know-who)	Methodology for mobilising individuals and organisations to take responsibility/ assume a particular role for implementing and supporting security action – comes under Involvement in 5Is.
Crime Role Grid (Know-who)	Tabulates crime roles x civil roles (e.g. offender x householder) to aid the targeting of offenders, involvement of crime preventers and blocking of crime promoters in complex crimes.
Conjunction of Criminal Opportunity and Conjunction of Terrorist Opportunity (Know-about and Know-what works)	Integrated, one-stop-shop framework of immediate causes of criminal events, and principles of intervention to block them, to aid analysis of crime problems and generation of theoretically-plausible interventions. Equivalent to Problem Analysis Triangle, but more detailed/extensive. The terrorist version is adapted to handle terrorism events.
Ds (Know-what works)	How security interventions work by influencing the offender, offering a fresh perspective for generating and evaluating action.
Disruption (Know-what works)	Diverse ways of adversely influencing groups/ networks of organised criminals/ terrorists – going beyond the customary superficial use of the term to stimulate better thought-through interventions.
Co-Eco-Devo-Evo (Know-about, Know-what, Know-how)	Interactions between offenders and security on a range of timeframes, from here-and-now, to learning/resource acquisition, to arms races, to facilitate better tactical and strategic planning of intervention.
Misdeeds & Security (Know-about and Know-what works)	How offenders may exploit or harm things and people in their environment, and how security can respond; an aid to forecasting which maps out the generic types of crime risk that may be associated with a particular kind of new technology or design of product, place, procedure or system; and the counterpart opportunities for prevention.
Security Function and Vibrant-Secure Function (Know-what and Know-how)	Systematic way of specifying security requirements for new designs, or describing security-related aspects of existing designs – purpose, security niche, mechanism, technicality. The vibrant version combines what we want more of with what we want less of.

Design specification for knowledge frameworks

A more formal design specification for knowledge frameworks includes:

• Appropriate complexity (discussed below) – neither too simple to handle the messy reality of security on the ground – nor too complex. Ashby’s Law of Requisite Variety essentially states that a framework must have a sufficient number of different internal states to be able to represent and manage the complexity of that part of the world out there that it is modelling. Paradoxically, investing in a little more effort to learn a more complex framework – if well-constituted and designed – can then simplify the effort for the user when using that framework in everyday practice.
• Conceptual precision and a controlled vocabulary; consistent, in-depth definitions for key terms, combined in integrated suites.
• Integration of theories in a common framework and language. The apparent simplicity of the individual core Crime Science perspectives (Routine Activities, Rational Choice etc) masks the fact that users have to find their own ways of handling overlapping concepts and gaps, as discussed in Ekblom (2006a, 2011c).
• Distinction between theoretical mechanisms and principles by which security interventions work, e.g. deterrence, vs practical methods, e.g. increasing the certainty of being caught by upgrading surveillance system.
• Use of tested theory and recombination of methods, to generate plausible innovations which can then be trialled and improved or rejected.
• A futures orientation – anticipation of new crime better than catch-up.
• Having ways of handling causal interactions, especially between intervention mechanisms and contextual variables; different ecological levels e.g. individual, family, organisation, community; and emergence.
• Attention to discourses used – humans as caused agents, both caused and causing.
• Articulation and management of often tacit practice knowledge for selection, replication and innovation in new contexts/for new problems.

Learning engines

Knowledge frameworks, once created, should not be fixed, once-and-for-all schemata. It is important that they are capable of adapting as they progressively get to know the nuances of their home range, as they make contact with wider domains and other disciplines each with their own frameworks, and as they keep up with changing theory, practice and research. Hence they should be designed as learning engines.

The development of the Crime Frameworks, has followed a process akin to the developmental psychologist Piaget’s concept of adaptive learning, where there is a dynamic balance between assimilation and accommodation – e.g. see a simple account here and an in-depth discussion here.

Assimilation is about fitting new information from the environment into one’s existing mental schema, and accommodation occurs when the existing schema has to change to fit the environment. A Crime Science example is the concept of target softening – typically, making products or buildings flexible and resilient rather than hard and resistant to attack.

One instance of assimilation versus accommodation is this padlock. It has a straight shackle which swivels in its housing so that it cannot be sawn through.

How to classify this lock within the 25 Techniques of Situational Crime Prevention? Functionally, it’s equivalent to target-hardening, but in terms of physical mechanism, it clearly differs. The challenge to the classifier is whether to squeeze it into the former, pre-existing category – assimilation – or create a new category of target-softening – accommodation. This example is of an existing framework adapting to something new; but the same challenge must be faced whenever we develop an entirely new framework, as with, say, the Conjunction of Criminal Opportunity or the Ds.

Appropriate complexity

The dilemma is that of determining the appropriate complexity of a given framework (also discussed above and elsewhere (Ekblom 2006a, 2007a, 2011c)). Are users of the framework to be better served by something that is simple but sometimes ill-fitting, versus a better match for reality, but requiring a more nuanced understanding and an initially greater learning effort?

The answer is rarely straightforward, but judgement calls must be made both tactically (how precisely do I deal with this instance?) and strategically (is this framework intended to be simple and robust versus elaborate and challenging?). These apply both during the early stages of developing a schema while the basic distinctions and concepts are being sketched out, and later when more refined polishing is in progress.

Of course, there is always the possibility that a well-developed framework is subsequently confronted by an instance that actually blows a major structural element apart, and the whole thing has to be thoroughly revised. (Particle physicists are waiting, some of them eagerly, for this to happen with their Standard Model.)

Making more advanced frameworks usable

Crime and security, as conflicting human interactions on many geographical and temporal scales and across all domains of life, are inherently complex. Societies are complex adaptive systems in which a diversity of active agents are each responding to, and perhaps anticipating, each other’s actions and to changes in their common environment. Mental schemas that enable people to handle this complexity have to be sufficiently detailed and sophisticated in themselves or they are part of the problem, not the solution. The need for this advanced approach is further explained in Ekblom (2006a, 2011c), on this page and in this presentation.

Because the Crime Frameworks do not oversimplify the complexities of the real world, users need extra support to learn and apply them.

There are various design techniques which can boost the usability of advanced frameworks. They can be introduced by gradually building the concepts and distinctions, rather than throwing users in at the deep end of the pool. Graphics can help those who respond better to visual representations. Interactive texts or graphics may be especially helpful, for example expanding/ collapsing headings or tree diagrams. Interactive toolkits, e.g. here, can lead users through the thinking/decision process whilst setting out an orderly sequence, limiting what needs to be considered at any one step, and creating navigation and audit trails for data logged and decisions made. And gamification can help train users in an engaging way – as with this example for teaching the Conjunction of Criminal Opportunity Framework.

Sometimes, though, too much precision can be a hindrance, especially when ideas are still developing; or when complexity is just too much to handle. Developing the 5Is framework, involved a conscious design decision to shift from an approach which was highly classificatory and set out descriptions of preventive activity via a rigid linear format, towards one that more closely resembled a language, which could be used to articulate action in whatever way users felt was helpful for their particular case and the use to which it was to be put.

Glossaries and ontologies

Knowledge can be formalised through the development of consistent systems of terms, definitions and concepts, known as glossaries and ontologies. They are discussed here, including an attempt at a visual glossary.

Publications and presentations

Publications

Presentations

Other publications also addressing knowledge issues

By other authors