How to Handle a Communication Crisis. An Epistemological Approach and a Neuro-Linguistic Contribution

  
 41
 
3.3 Unknown prodromes 
The third type of prodromes includes all those warning signs that are specific to a particular 
organization. Each organization has its own history, culture, physiognomy, internal dynamics, 
business model, network of stakeholders, etc. These specificities may origin unorthodox crises that 
are preannounced by bizarre warning signs. Actually, as already said in §1.1, each crisis, being an 
event, is unique by nature; however it is often possible to include a real crisis into a formalized class 
of crises that share common characteristics and, by so doing, cope with them by following some 
pre-determined paths, methods, procedures, etc. 
When a crisis is completely peculiar to one organization, it is almost impossible to recognize any 
kind of prodromes or warning signs, in this case the organization can avoid the crisis only by 
relying on the intuition of management or (if the lines of communication permit it) by the 
qualification and commitment of employees. Of course not having any kind of benchmark will 
negatively affect the possibilities to detect the warning signs and to avoid the crisis. 
  
 42
 
3.4 Prodromes of a crisis or simple noise?  
Finally, and in cauda venenum, the fourth kind of prodromes: the most difficult to detect. There are 
signals that are warning signs and signals that don’t preannounce anything, that are just noise. I call 
it junk information.  
The problem is that, formally, there may be no difference between a prodrome and a signal that is 
exactly like a prodrome but that actually is not the indicator of an imminent crisis.  
Only an analysis a posteriori can determine if a signal is the sign of something and so has to be 
taken into consideration or if it is just noise that, if taken into consideration, will diminish the 
efficiency of the organization until it totally compromises the efficacy of the system and sometimes 
jeopardizes the achievement of the organization’s goal. 
Critics of an optimistic approach to prevent crises by monitoring and detecting warning signs have 
an easy task in dismantling this optimism, as highlighted by Hopkins:  
“The warning signs are only obvious in retrospect and it is often not possible to discern their 
significance beforehand. The point is often put in terms of the signal/noise metaphor. For instance, 
Perrow has argued that although there were warnings prior to the near disaster at Three Mile 
Island nuclear power station un 1979, it would have been impossible to distinguish signal from 
noise beforehand” (Hopkins, 2007, p.7).  
According to Perrow in fact: “Signals are simply viewed as background noise until their meaning is 
disclosed by an accident” (Perrow, 1982, p.175) 
The heart of the matter is a confusion between the relations of simple association and causal 
inference that I’m going to examine in the next paragraph. 
  
 43
 
3.5 The Problem of Causation 
To continue the metaphor begun in §3.4, it was suggested that latent failures were analogous to the 
“resident pathogens” within the human body, which combine with external factors (stress, toxic 
agencies, etc.) to bring about disease (Cfr. Reason in Smith & Elliott, 2006, p. 249). Like cancers 
and cardiovascular disorders, accidents do not arise from single causes, they occur through the 
unforeseen concatenation of several distinct factors, each one necessary but singly insufficient to 
cause the crisis (Ibidem).  
Now, in order to detect the warning signs of a crisis the possibility of understanding what causes 
what is taken for granted  
Each event is caused by a long, perhaps infinite, list of causes. 
Try to imagine how many causes it is possible to find before establishing what caused a football 
match to be won by, say, Juventus F.C. with a penalty transformed into a goal by Del Piero. The 
ability of Del Piero to kick the penalty. The inability of the goalkeeper to save the goal. The ability 
of Trezeguet to force the opponent defence to make a fault within the penalty area. The inability of 
the defence to fairly stop Trezeguet. The referee who always is on Juventus’s side. And more: the 
coaches, the supporters, the weather conditions, the football pitch, etc. etc. The readers can imagine 
thousands of other factors which determined that exact situation. This is a way to explain why 
supporters are able to discuss a match for hours without reaching a consensual agree on what 
happened. Actually, innumerable things happened! And each one had its own impact (even if 
remote) on the event to explain. 
The law (Italian and International) also had to face the problem of causal attribution, in order to 
determine who is responsible for an event and to which extent
37
.  
Generally the causes are distinguished between direct and indirect. In a legal litigation this 
fundamentum divisionis is enough to attribute a fault to a part or to another, of course it is not an 
exact science rather it is based on a human convention but since litigations are a human matter this 
method of attribution works quite well.  
The situation changes when dealing with crises. From a posteriori it is quite easy to charge 
responsibilities of a crisis to one or the other factor (in the same way ordinary justice acts), but if 
our intent is to prevent a crisis, then it is no longer enough to establish which was the ultimate cause, 
we should take into consideration the entire causal chain.  
A brief preamble is here required. 
First of all we have to avoid the logical fallacy of “cum hoc ergo propter hoc” (With this, therefore 
because of this), in statistical sciences known with the statement “Correlation is not Causation”
38
.  
                                                 
37
 See for example the IPSOA dossier (2005) 
  
 44
                                                                                                                                                                 
This misinterpretation of causality leads to paradoxes such as: 
¾ There is a high correlation between the number of fire-fighters and the entity of the damage 
provoked by the fire. Do the fire-fighters make the situation worse?  
¾ A very high percentage of people die in a bed. Is the bed a dangerous place in which to stay?  
¾ The areas in which the number of storks is higher are also the areas in which it records a 
high birth rate. Do the storks bring the children
39
?  
Fisher brought in another eloquent example: “Near-perfect correlations exist between the death rate 
in Hyderabad, India, from 1911 to 1919, and variations in the membership of the International 
Association of Machinists during the same period. Nobody seriously believes that there is anything 
more than a coincidence in that odd and insignificant fact”. (Fisher, 1970, p. 168).  
 
To a certain extent, the problem of causation can also be represented in a formalized way through 
the regression equations specified in 1964 by Blalock with the following formulation: Y
i
 = a + 
bX
i
 + e
i
Being: 
- Y the outcome variable of interest (Explanandum);  
- X the causal variable of interest (Explanans);  
- a and b the regression coefficients (a indicates where the regression line intersects the Y 
axis and b the slope of the regression line);  
- and finally e is the approximation error.  
The acute question raised by Blalock is the following: 
“What if there existed a major determinant of Y
i
, not explicitly contained in the regression equation, 
which was in fact correlated with some of the independent variables X
i
? Clearly, it would be 
contributing to the error term in a manner so as to make the errors systematically related to these 
particular X
i
. If we were in a position to bring this unknown variable into the regression equation, 
we would find that at least some of the regression coefficients (slopes) would be changed. This is 
obviously an unsatisfactory state of affairs, making it nearly impossible to state accurate scientific 
generalizations”. (Blalock, 1964, p. 47).   
However, the most important and complete theory of explanation in Social Sciences is the one 
provided by Hempel and Oppenheim at the end of the 1940s; after it, several other theories have 
appeared but the hard core of the Hempel and Oppenheim’s Theory is still to be outpaced.  
 
38
 The first philosopher who clearly raises the question is David Hume. According to the Scottish thinker: “We have no 
"necessary connection" between the sense data and the proposition of causation: we see merely one thing happening 
and the other thing following and therefore we often presume (without knowing), that one is causing the other, but it is 
perfectly possible not to presume it. (…)The first time one sees this [a correlation between two events], they are merely 
"conjoined" events; but as it repeats itself over time, imagination takes over and we begin to ascribe a "connection" - 
but this is only a human "feeling", that when we say X causes Y, we mean only that they have acquired a connexion in 
our thought and give rise to this inference [of causation]." (Hume, 1748,  p.82) 
39
 In the first two cases the explication is straightforward, in the third “paradox” there is an hidden variable which is 
directly connected with the number of births and with the high presence of storks: a rural area. People in a rural area 
tend to have more babies and a rural ambient is also the natural context of storks. For a deeper insight see Rosenberg, 
1968, pp. 54-83.