Friday, April 25, 2014

Occam vs. Einstein - Crime science gets a facelift

Achilles and the heel - photo Creative Commons by Tasoskessaris 

[SPOILER ALERT - This is a rather long theory blog on emerging research. Apologies. But every now and then, like bad tasting medicine, it's necessary to ingest!]

Yesterday I was external examiner for an MA thesis defense on CPTED and the geography of youth gun violence. The candidate did a great job. She passed her defense and is now a new criminology scholar. Reading her literature review I was struck how difficult it is for new scholars to siphon out decent research. Crime and place studies range from sensible to silly.

Not long ago I got to do some siphoning of my own – for better or worse - while reading two crime studies. The better version was titled “The reasoning criminal vs. Homer Simpson: conceptual challenges for crime science”. The worse version was torturous!

A CPTED Framework Public Realm Scoping Paper was released two years ago. While it did an admirable job of covering CPTED history, particularly the Australian story, it read like exotic buffet where each tasty morsel was spread out without rhyme or reason.


Framework tossed out ideas like "thermal comfort" and "empathic interaction" in willy nilly fashion. On one hand it slammed 2nd Generation CPTED theory as “superfluous second generation nomenclature” and then followed that up with this poster-child for superfluous:

“Whole area image, in a Gestalt sense, is very different from the meaning embedded in individual site characteristics… Past psycho-social experiences, role-models, somatic and genetic tendencies and inheritance, introvert-extroversion personality-typing, psychological stressor thresholds, 'get even' desires, thrill seeking, peer pressures, and gang membership, inter alia...encourage individuals considering a delinquent, anti-social or criminal activity to take action (or not)."


Ironically the author then claims social programming has been part of CPTED all along. It hasn't! That’s why we created 2nd Generation CPTED in the first place.

Into this theoretical bucket the author scoops environmental, thermal, planning, and design cases as CPTED examples, all the while supporting bottom-up ideas similar to SafeGrowth ("local control is best practice"). But then we’re offered up cases such as a Viennese coffeehouse that "employs people suffering from mental handicaps…aiming to stabilize the personality of the participants."


Framework is a frustrating read. There is a rich historical review in this paper. The author is clearly knowledgeable (in the 1990s he led pioneering research into fear mapping). Sadly this is not his best work. It is in dire need of a disciplined copy editor.

As luck would have it my next read was about Homer Simpson in an article about the emerging field of crime science!


Last October Frontiers in Human Neuroscience published Naemie Bouhana's article, The reasoning criminal vs. Homer Simpson: Conceptual challenges for crime science.

It was concise, non-pretentious and penetrating. It described a new branch of criminology – crime science – as the study of crime prevention "chiefly concerned with the design of social and technological systems…an engineering discipline, with a self-confessed preference for short-term problem-solving."

I suspect crime science researchers see their role a bit broader. Last year their inaugural journal said crime science is about cutting crime opportunities. “In blocking opportunities for crime and terrorism we are not simply reducing the incidence, we are also removing one of the causes.”

Obviously CPTED is a practical piece that fits neatly into the center of that puzzle. Yet crime science has an Achilles Heel: "It is not possible to leave the offender out of crime prevention altogether. In order to “increase effort” and “reduce rewards”, a model of criminal decision-making is needed. For this purpose, the fathers of situational crime prevention adopted the Rational Choice Perspective (RCP).”

Bouhana claims that RCP "has fallen short as a model of offender decision-making." He explains why with a razor.


Occam's Razor says theories with the fewest assumptions are preferable to those more complex. Aristole said it first: all things being equal, theories that explain the world with fewer hypotheses are better. Science calls it parsimony.

The parsimony razor helps shave away the complex to arrive at the simplest solution. RCP is the perfect theory for crime reduction and opportunity the perfect tool cut it; cut the opportunity and you cut the crime. It is Occam’s Razor incarnate.

Bouhana says there are three arrows through the heel of RCP in crime science:

  1. In strictly scientific terms, parsimony is never irrefutable. Especially in modern science, complexity often works best. RCP’s simplicity may be its weakness, not its strength.
  2. When are all things ever equal in the social world? 
  3. Einstein warned that a theory should never simplify to the point of sacrificing "the adequate representation of a single datum of experience." In other words, reducing prevention to “where” and “when” leaves out too many of the big questions, like “why”. 

In criminal behavior people don’t always act like Mr. Spock - logical and rational. Bouhana says crime science must expand to include new theories in neuroscience and behavioral geography. Those theories suggest people often behave irrationally, more like Homer Simpson, sometimes responding to the environment, sometimes not. Consider the well-known "By-stander Effect"; natural surveillance did nothing to stop the Kitty Genovese murder in New York.

"This state of affairs has had the consequence of stifling theoretical development in crime science, so much so that RCP has remained essentially static since the 1980s." This is exactly what many of us have said about 1st Generation CPTED for years.

In addition to neuroscience and behavioral geography, I’ve mentioned other theories in this blog that crime science might consider like emotional intelligence, the civilizing effect, and the public health concept.

These new theories are not parsimony. Like the criminal behavior they study, they are complex. Occam would not approve, though Einstein might.

4 Replies so far - Add your comment

Jon Lusher said...

Greg, I understand the literary use of simile, metaphor and the like, but I must take exception to the Einstein part of this usage. E=mc2 is about as close to the bone Occam would prefer as one can get. Even the explanation of relativity is a picture-thought of extreme simplicity; just because we mere mortals can't grasp it isn't because it's a too-complicated theorem. Anyway, your points are well taken, and the gibberish expressed in pseudo-scientific language is a cover for poor science and even poorer explanation.

GSaville said...

Thanks Jon for that fun twist on Einstein and my metaphors. In half-hearted defense, I took many of those metaphors (Spock, Homer Simpson) directly from Bouhana's article. And of course Occam arises directly from parsimony - which was the whole point.

The Einstein quote was from, obviously, Einstein himself as he talked about the limits to parsimony. I do suspect though his actual theory is much more complicated than the well-known formula implies - it's probably, as you say, mere mortals like us have no idea what it really means.

Tim Hegarty said...

The world is complex, and it order to make some sense of it, we employ frameworks that help to simplify it for us. RCP is a framework to help us understand some aspects of crime, just as the By-Stander Effect is a framework to help us understand why people don't act. Yes, frameworks generalize, and they over-simplify at times, but without them, I think we would be paralyzed in the face of complex problems. I think the bigger issue is a tendency to rely on only one framework. When all you have is a hammer, every problem looks like a nail.

By the way, pop cultural metaphors are another way to help make sense of the complex. I grok Spock.

GSaville said...

Tim, there are probably on 6 nerdy geeks in the whole world who actually know the meaning of "I grok Spock". Sadly, I am one of them.

Thanks for the insightful comments.

By the way, frameworks to understand the world don't always have to be simple - they also can be complex like the world they describe. Those too are helpful!