Expansion of thinking about "cycle of abuse:" General considerations

["Thomas R. Chibucos" <tchibuc@xxxxxxxxxxxxxx>]

Mark Chaffin makes some very clear and salient points about how an early
event or marker can serve as a risk factor without necessarily (or usually,
I add) producing a simple correlation with some outcome later in time--and,
of course, without implying a causal relationship.  His argument is
important because it really can apply to ALL HUMAN BEHAVIOR, and virtually
all outcomes--whether they are highly problematic or whether they are more
positive.

Consider the following general theoretical approaches and their underlying
principles.

1) life-span developmental theory and research--essentially that there is
both continuity and change across time depending on the quantity, quality,
and interaction of numerous variables;

2) the ecological and systemic conceptualization of development (of
individuals, of families, of social units)--essentially that any given
behavior depends to one degree or another on the nich(es) in which it is
embedded; and

3) 'developmental contextualism'--essentially that the RELATIONSHIPS between
changing individuals and changing contexts is key to understanding human
behavioral development.

In the relatively young science of human behavior (approx 100 years old or
so) one of very few strong laws seems to be:  The longer and the more
intense a variable operates on a person, and given an often large and
bewildering array of intervening and reciprocally interacting variables, the
more likely that the variable will correlate with some outcome.  However, as
Chaffin notes this does not necessarily imply simple linear causality
between X at time point 1 and Y at time point 2.  There are a variety of
conceptual and substantive reasons that demand caution in interpretation, as
indicated by the above 3 listed theoretical approaches and principles.  But,
there are also methodological and statistical reasons for such
caution--e.g., the closer in time to one another that two events that are
being correlated are, the less likely they are to be independent.  This
autocorrelation becomes less of a problem the greater the time interval
between correlated variables.

We are at the beginning (or is it well into?) a more sophisticated era of
contrary thinking, however.  Mapping the human genome will produce,
disguised as real understanding, yet another spasm of simplistic linear
thinking about causal influences.  But, just as the temporal succession of
events, variables, etc. following abuse is clearly influential in relation
to a subsequently measured outcome so does the genotype (in the species and
in an individual life) "develop" in its own ecologies.

As a not too attractive but very smart man once said to a very beautiful
woman of limited mental acuity after she suggested that they could make the
perfect child together: 'I would not be so sure, my dear.  The child could
be born with my looks and your brains."  Although an admittedly somewhat
crude statement, this perspective should form the foundation for
interpretation of ALL human development research and practice, whether the
focus is on specific behavioral or genetic antecedents.


-----Original Message-----
From: owner-CHILD-MALTREATMENT-RESEARCH-L@cornell.edu
[mailto:owner-CHILD-MALTREATMENT-RESEARCH-L@cornell.edu]On Behalf Of
Mark J. Chaffin
Sent: Thursday, October 12, 2000 10:37 AM
To: Child Maltreatment Researchers
Subject: cycle of abuse


Julie,

I don't think the main weakness of the 'cycle of abuse' theory is based on
the idea that not all abused children grow to become abusers or criminals,
or upon occasional exceptions to the 'cycle' (although the exceptions are
actually the rule, not exceptions--its the vast majority of abused children
who grow up to be prosocial and nonabusive).  Risk is, of course, a
probabilistic construct.  However, I think the main weakness of the 'cycle
of violence' theory is that its a bit like the cocaine exposed baby theory.
The theory is that cocaine exposure in utereo causes profound behavioral (or
criminal) problems later in life.  And you can certainly find data to
support a simple correlational relationship, just as you can with the 'cycle
of abuse' theory.  The problem is that cocaine exposed babies usually grow
up with drug addict parents, living in stark poverty and disintegrating
neighborhoods, attend marginalized schools, witness and experience violence,
etc. etc.  These are parts of a whole.  Research efforts to separate these
influences have suggested that it may be the environmental influences which
have a larger impact than the drug exposure.  So, to complete the analogy,
its not just a question of whether abuse history is a probabilistic
predictive (or risk) factor.  Its a question of how much abuse history is a
causative factor, and how much its a marker variable for other things.

My impression from the admittedly incomplete prospective data on the
question is that abuse history is important, but when one considers the
larger ecological picture, it may not be as important as other factors in
understanding juvenile violence.


Mark Chaffin, Ph.D.
Center on Child Abuse and Neglect
University of Oklahoma Health Sciences Center
P.O. Box 26901; CHO 3406
Oklahoma City, OK  73190
(405) 271-8858; fax 271-2931
mark-chaffin@ouhsc.edu