Structuring the Mess

Individual views of the stakeholders on the problem situation, recorded in the minutes of interviews with them, are the initial data containing information about the system under consideration, still in an implicit form.

Extracting information is expressed in the construction of more and more meaningful models (data, information, knowledge, understanding, wisdom — see Chapter 3, Part I).

The first step is to sort, group, and classify data (results of observations and measurements). This initial processing gives a description, called by R. Ackoff information. Considering the stage of the formation of a mess, this operation leads to the formation of groups of problems in the static description of the mess and to the emergence of scenarios (reference projections) in the dynamic one.

There are technologies that do not require further structuring of the problematic mess (e.g., idealized design). However, there may be a situation when it is necessary to proceed to considering the interrelations between the components of the problematic mess, and to build models of the next level, such as “knowledge”.

A typical example of this is the situation when it is clear at the start that the available resources are not enough to solve all the problems that make up the mess.

There are questions about which problems to give preference and how to take into account all the others correctly. The answer is to design a sequence of improving interventions for different resource allocations for mess problems, and from all such options, to choose the one that gives the greatest improvement under the given constraints (i.e., optimal). Usually, among the limitations, in addition to the volume of resources, there is a mandatory improvement in the situation for the client, but at the same time, it may turn out that the main resources do not have to be directed to personal problems.

The task of allocating resources and determining the priority of solving problems is greatly facilitated if we manage to somehow sort the mess. However, this is not always possible. An example, when possible, is planning a targeted project, in which each participant must identify problems for themselves and fulfill their role. In this case, after identifying the mess, it is recommended to start building a hierarchical problem tree, thereby revealing the cause-effect relationships between them.

One variant of the problem tree construction technique can be described as follows. Stakeholders (project participants), having familiarized themselves with the mess, write out, at their own discretion, a key, focal problem in the entire mess. Everyone is guided by their own interests and their problems. Then there is a collective discussion of the ranking of problems until the participants agree on which problem is the starting one. Subsequently, the next problem is taken and compared with the first one. Then:

• - if the second problem is the cause, the condition for the first, it is placed a level below;
• - if it is a consequence of the first, it is placed higher;
• - if it is neither a consequence nor a cause, then it is placed on the same level.

This scheme addresses the remaining problems. In the course of work, it may be appropriate to change the focal problem, but this does not deprive the meaning and significance of the analysis.

For example, if the focal problem is “Insufficient number of qualified professionals in such and such a specialty”, then its reason can be formulated as “Insufficient level and quality of higher education in this specialty”.

In the presence of a problem tree, the question of the sequence of problems is solved in an obvious way: the higher problems cannot be solved until the lower ones are solved.

However, cases where problems can be built into an ideal tree-like hierarchical structure are rare. Because of the general interconnection in nature (a consequence of the second property of the system), more often, only the essential connections between the components of the mess form a network-like structure, in which the mutual influences sometimes form closed feedback loops. This is what leads to the understanding that for complex problems there are no simple solutions, and that any problem has no single reason, although among them there are “more important” and “less important” ones.

In general, even when the mess is not tree-like perfectly, the sequence and intensity of problem-solving should be determined precisely by the structure of the mess, the desire to move the whole mess to improvement, and not to those of the stakeholders who are more persistent in seeking to solve their own problems.

J. Warfield [3] proposes a different approach to structuring a problem mess, that is, to build not a tree-like but a network structure. He calls this “explanatory model building” (ISM, interpretive structural modeling).

For each pair of problems, the question is: does the problem “x” exacerbate the problem “y”? The answer is the basis for including the edge between “x” and “y” in the graph and applying a direction to the edge. The wording “x” and “y” may need to be clarified. The decision on the existence of a relationship between “x” and “y” and its focus was adopted by a vote of the majority (“don’t know” equates to “no”). Consequently, we obtain a directed graph describing the structure of the mess.

For a large number of pairs in question, this work can be done with the help of computers (the corresponding programs can be downloaded free of charge from http://www.gmu.edu/depts/t-iasis/).

The resulting graph can be further transformed into a smaller number of simpler second- and third-order coarsened structures, grouping the problems by some features. For example, the author recommends to group the problems according to their importance. Each stakeholder is invited to choose from five major problems and continue to work with them, for example, analyzing their overlap.

The ultimate goal of structuring is to transform the graph so that the arrows are directed from left to right, highlighting the problems that should be solved first and those that can be solved later. Of course, there are difficulties in the case of cycles in the graph, which are subject to separate consideration.

Structuring the mess is a preparatory step to a consistent solution of a set of all problems.

However, the method of successive, small, and “continuous improvement” is criticized by many as a half-measure, poorly coordinated with the variability of the environment, without fully taking into account the emergence of the system.

When we remove the unwanted, we often obtain even worse one. Getting rid of a television program by switching to another channel, we get to a program that we like even less. The use of dust against harmful insects was harmful to the environment. The legal prohibition of alcohol gave rise to organized crime. Prisoners in jail often only increase their criminal skills. Therefore, getting rid of the undesirable is often just a departure from responsibility.

As an alternative to the incremental approach, the method of “idealized design” by R. Ackoff deserves special mention. This technique does not require the ordering of the problem mess: the identification of problems and their consequences are only needed to understand the need for change. The subsequent design of changes comes from the assumption that the existing (and not satisfying us) system “disappeared last night”, and the work is not aimed at correcting undesirable features of the organization, but at the design of the desired.

Participation of Stakeholders in the Analysis

The best source of reliable, accurate, and complete information about the stakeholder is, of course, he himself. To find out what his problems are, you need to contact him and talk to him. So we faced the need to involve stakeholders in the process of analysis. (It will later become clear that this is necessary for the entire process, not just at this stage.)

No one can express their opinion better than they do. But getting first-hand information faces the problem of stakeholder availability. “Silent” stakeholders are by nature noncontact; some participants in the situation do not have the opportunity or desire to cooperate with the analyst; some stakeholders may not be a person, but a group, and even numerous; and some may not be personally available for geographical or political reasons. But information about them is still needed!

The problem of inaccessibility can be solved as follows. All stakeholders are divided into two types: obligatory and desirable participants of the analysis.

First, you must determine problem-keeping and problem-solving systems among organization’s stakeholders. In the problem-holder system, the question is clear: it is the representative who asked the analyst to become his client. Next, it is necessary to identify the problem-solving system(s), that is, those at whose expense and with whose help the intervention will be implemented. Often the same organization is both problem-holding and problem-solving (e.g., a powerful firm), but sometimes they differ (e.g., school and the city council, municipal and federal authorities).

It is of fundamental importance that the first persons of the problem-holding and problem-solving systems are obligatory participants of the system analysis. Without their participation, all the work has no practical meaning, which makes it hopeless: after all, they have the resources and powers necessary to solve the problem. That is why the guarantee of their participation is included in the contract, and if this condition is not accepted, the analyst does not take up work.

Other stakeholders are desirable participants. That is, it is best if they can be included in the collective process of systems analysis. But if someone is inaccessible, it is possible to obtain the desired information about him and without his personal participation. This possibility is based on the consequence of the second property of the system — the universal relationship in nature. There is a chain of subjects between the analyst and the inaccessible stakeholder. Being related, they contain information about each other, the greater the closer they are to each other. Therefore, we can try to obtain information about the inaccessible stakeholder to get in touch with those nearest to him in the chain and “extract” the information we need from them instead. This, of course, is less reliable than the stakeholder, the source of information about it, but we have no better opportunity. The last thing is for the analyst to express the opinion of the stakeholder: he is the most distant link in this chain. This is only allowed if everyone else in the chain is unavailable.

The task of finding a sufficiently knowledgeable representative of an inaccessible stakeholder is not as difficult as it seems. Probably everyone had to talk to a stranger to find that you have common friends. There was even a scientific study of this issue by scientists at the University of California. The target person was represented by a student of this university. Then, from the list of voters of the most remote from California, Pennsylvania, was randomly selected 100 people who agreed to participate in the experiment. Each of them was given a letter addressed to the girl, but it was possible to send it directly to her only on condition that the sender knows her personally. If not, the letter was sent to one of his personal friends on the same condition. The U.S. mail provides a service of correspondence through intermediaries, so that the letters eventually reached the addressee. The length of a particular chain was determined by the number of post stamps. It turned out that in order for two strangers out of 200 million people contacted by a chain of people who know each other, you need quite a few links. The shortest chain was two (a farmer has his childhood friend a Congressman in Washington, and the Congressman’s friend is a professor at that university), the longest was 11, and the average length was 4.5. It should not be surprising that with an average of 100 acquaintances, a network of chains of four people will cover 100 million individuals.

So, in case of unavailability of the stakeholders, it is necessary to find their best possible representatives and involve them in the team working on the solution of the problem. In particular, representatives of “silent” stakeholders can be experts, scientists, heads of relevant administrative bodies, and public organizations of relevant profiles.

It remains to say that if the stakeholder is a group of people (students of the city, pensioners, entrepreneurs, etc.), then either try to attract a competent representative of the group (among its leaders), or resort to the methods of applied sociology and statistics (to clarify public opinion).

Hence, we assume that the selected, instructed, and inspired by the systems analyst group of stakeholders (or their competent representatives) formulated the mess and structured it as possible.

Questions and Tasks

• 1. What do you call a “problem mess”?
• 2. What is the dynamic version of the problem mess?
• 3. Why not solve the problem of the client in isolation from the problem mess?
• 4. What does it mean to “work with a mess as a whole”?
• 5. How to solve the difficulties arising from inaccessibility of the stakeholders?