Archive for February 2012
Decision-making is a key activity at all levels in an organisation. All employees make decisions: from the front-line employee who has to decide how to handle a difficult customer to an executive who has to choose between projects that are competing for funding. Given this it is no surprise that a vast body of knowledge – decision theory – has been developed to support the process of decision making.
Decision theory concerns itself with rational decision making– that is, decisions that are based on an objective evaluation of available options and their consequences, leading to a choice that is made on the basis of such an evaluation alone. In reality, though, many decisions are not made this way. In this post I look at the different ways in which decisions are actually made in organisations, drawing on a brilliant essay by James March entitled, How Decisions Happen in Organizations.
Decision making as a rational process
The standard view of decision making is that it is a process of rational choice based on:
- Knowledge of alternatives
- Knowledge of the consequences of each of the alternatives
- Ordered preferences by which consequences can be evaluated
- Rule(s) by which a particular alternative can be selected
In its basic form, decision theory assumes that each of the above is fully known. As March states:
In the most familiar form of the model, we assume that all alternatives, the probability distribution of consequences conditional on each alternative, and the subjective value of each possible consequence are known; we assume a choice is made by selecting the alternative with the highest expected value. This emphasis on expected value may be moderated by a risk preference (i.e.,some value associated with the variability of the outcome distribution).
However, there are a number of challenges to this ideal picture of decision-making. These include:
- Uncertainty about consequences of actions: The standard theory of rational choice assumes that decision-makers have knowledge of all possible outcomes of actions. However, this is not possible because humans are boundedly rational – their ability to seek and process information is limited by their cognitive abilities and available resources. Quite often it happens that consequences reveal themselves only after a decision has been made and implemented. As March mentions, “…management requires tolerance of the idea that the meaning of yesterday’s action will be discovered in the experiences and interpretations of today…”
- Uncertainty about preferences: The standard theory assumes that preferences are stable and consistent. Quite often, it happens that preferences change with time and different preferences can be inconsistent with each other.
- The role of risk: Typically, in theories of rational decision making risk appetite (of an individual or organisation) is treated as a single fixed number. In reality, it varies with situational factors such as level of threat to survival, excess resources available etc. Moreover, it also depends on the (often unarticulated) hopes and fears of individuals who are making the decision.
- Conflict between decision makers: Rational theories of decision making assume that conflict between decision makers can be resolved by (rationally!) evaluating conflicting alternatives and choosing the best one based on an agreed decision rule. The problem is that in such situations it is often impossible to come up with such a decision rule. Negotiations over criteria can go on interminably and conclude without agreement. March suggests that the reasons why decisions get made despite this is that people rely on trust and reputation rather than formal agreements in order to reach a consensus.
So as we see, the rational view of decision making has less practical relevance than one might expect. It is part of the story of decision making, but definitely not the whole tale.
Decision making as a rule-based activity
An alternate logic of decision making is that of following rules, obligations and duties; doing what is appropriate rather than what is rational. As March puts it:
Much of the decision-making behavior we observe reflects the routine way in which people do what they are supposed to do. For example, most of the time, the majority of people in organizations follow rules, even when it is not obviously in their self-interest to do so. Much of the behavior in an organization is specified by standard operating procedures, professional standards, cultural norms, and institutional structures. The terminology is one of duties and roles rather than anticipatory, consequential choice.
Within a logic of appropriateness, people make decisions by mapping the aspects of the decision they are required to make to what is appropriate in such situations. In particular, they consider the following:
- Situation: what kind of a situation is this?
- Identity: who am I? What kind of position do I hold in the organisation?
- Determining an appropriate choice: What should a person like me (or in my position) do in this kind of situation?
In such a process the focus is on doing what is right (as per the rules) rather than searching for rationally determined best choice.
The interesting question is how these rules come into existence. March describes three ways in which this happens:
- Rules are developed through experience and are modified by feedback on what worked well and what didn’t. In this view organisations create rules.
- Rules are selected (rather than developed) based on their suitability for a group or organisation. In this view, rules have an existence independent of organisations.
- Rules spread from organisation to organisation – much like “fads or measles.” In this view, rules are created in idiosyncratic ways (through an innovative or quirky choices made by an individual, say) and then, if they are successful, are copied others. Many popular management practices have their roots in such fads.
Summarising: decisions can be based on appropriate choices rather than rational ones.
Decision making as a contingent event
The views of decision making embodied in the logic of rationality and appropriateness assume that the cause-effect relationships between decisions and outcomes are well understood and that organizational rules and hierarchies actually control outcomes. However, in reality things tend to be less straightforward. For example:
- Many things happen at the same time, each competing for the attention of decision makers. The attention a decision maker gives to a problem thus depends on the other things that are on her mind at the time.
- Individual perceptions of situations vary, thus making the formulation of a decision problem difficult (in effect, making it a wicked problem).
In cases such as these, the decisions are contingent on factors that have nothing to do with the decision itself. As examples, an executive who is distracted by personal problems may not give enough attention to a decision problem at hand and a bunch of stakeholders who cannot agree may end up making a decision that cannot be justified via rationality or appropriateness.
Decision making as a byproduct of other factors
The assumption underlying the foregoing discussion is that decisions affect outcomes and hence that decisions matter. However, as March points out:
Descriptions of decision arenas often seem to make little sense in such terms. Information that is ostensibly gathered for decisions is often ignored. Contentiousness of the policies of an organization is often followed by apparent indifference about their implementation. Individuals fight for the right to participate in decision processes, but then do not exercise the right. Studies of managers consistently indicate that very little time is spent making decisions. Rather, managers seem to spend time meeting people and executing managerial performances.
Based on the above, March makes the interesting point decision making is often a ritual activity that has little to do with the actual decision itself. The process of making a choice provides decision makers opportunities to do other things such as:
- Presenting and justifying their viewpoints to their peers.
- Distributing credit or blame for what has occurred.
- Reaffirming loyalties and friendships
An aspect of decision making made highlighted in the previous section is that there are many competing demands on a decision maker’s attention – for example, family, friends or personal goals. This is true in general: in the course of our lives, we are presented with a steady stream of choices, opportunities and problems. The degree to which each of these hold our attention depends on a host of factors including (but not restricted to) our values, duties and priorities. Because of these concurrent or nearly concurrent issues, the attention we give to a decision problem is closely linked to events that have recently occurred or are anticipated in the near future, and the priorities we assign to these. In such situations, the logic of decision making is temporal (dictated by time) rather than consequential or rule-based. In other words, our decisions depend on recent events and our immediate (or recent) environment.
In this article I have summarised various views of decision making drawing on the work of James March. We have seen that the official line about decision making being a rational process that is concerned with optimizing choices on the basis of consequences and preferences is not the whole story. Our decisions are influenced by a host of other factors, ranging from the rules that govern our work lives to our desires and fears, or even what happened at home yesterday. In short: the choices we make often depend on things we are only dimly aware of.
Knowledge management (KM) is essentially about capturing and disseminating the know-how, insights and experiences that exist within an organisation. Although much is expected of KM initiatives, most end up delivering document repositories that are of as much help in managing knowledge as a bus is in getting to the moon. In this post I look into the question of why KM initiatives fail, drawing on a couple of sources that explore the personal nature of knowledge.
Explicit and tacit knowledge in KM
Most KM professionals are familiar with terms explicit and tacit knowledge. The first term refers to knowledge that can be expressed in writing or speech whereas the second refers to that which cannot. Examples of the former include driving directions (how to get from A to B) or a musical score; examples of the latter include the ability to drive or to play a musical instrument. This seems reasonable enough: a musician can learn how to play a piece by studying a score however a non-musician cannot learn to play an instrument by reading a book.
In their influential book, The Knowledge-Creating Company, Ikujiro Nonaka and Hirotaka Takeuchi proposed a model of knowledge creation1 based on their claim that: “human knowledge is created and expanded through social interaction between tacit knowledge and explicit knowledge.” It would take me too far afield to discuss their knowledge creation model in full here – see this article for a quick summary. However, the following aspects of it are relevant to the present discussion:
- The two forms of knowledge (tacit and explicit) can be converted from one to the other. In particular, it is possible to convert tacit knowledge to an explicit form.
- Knowledge can be transferred (from person to person).
In the remainder of this article I’ll discuss why these claims aren’t entirely valid.
All knowledge has tacit and explicit elements
In a paper entitled, Do we really understand tacit knowledge, Haridimos Tsoukas discusses why Nonaka and Takeuchi’s view of knowledge is incomplete, if not incorrect. To do so, he draws upon writings of the philosopher Michael Polanyi.
According to Polanyi, all knowledge has tacit and explicit elements. This is true even of theoretical knowledge that can be codified in symbols (mathematical knowledge, for example). Quoting from Tsoukas’ paper:
…if one takes a closer look at how theoretical (or codified) knowledge is actually used in practice, one will see the extent to which theoretical knowledge itself, far from being as objective, self-sustaining, and explicit as it is often taken to be, it is actually grounded on personal judgements and tacit commitments. Even the most theoretical form of knowledge, such as pure mathematics, cannot be a completely formalised system, since it is based for its application and development on the skills of mathematicians and how such skills are used in practice.
Mathematical proofs are written in a notation that is (supposed to be) completely unambiguous. Yet every mathematician will understand a proof (in the sense of its implications rather than its veracity) in his or her own way. Moreover, based on their personal understandings, some mathematicians will be able to derive insights that others won’t. Indeed this is how we distinguish between skilled and less skilled mathematicians.
Polanyi claimed that all knowing consists at least in part of skillful action because the knower participates in the act of understanding and assimilating what is known.
Lest this example seem too academic, let’s consider a more commonplace one taken from Tsoukas’ paper: that of a person reading a map.
Although a map is an explicit representation of location, in order to actually use a map to get from A to B a person needs to:
- Locate A on the map.
- Plot out a route from A to B.
- Traverse the plotted route by identifying landmarks, street names etc. in the real world and interpreting them in terms of the plotted route.
In other words, the person has to make use of his or her senses and cognitive abilities in order to use the (explicit) knowledge captured in the map. The point is that the person will do this in a way that he or she cannot fully explain to anyone else. In this sense, the person’s understanding (or knowledge) of what’s in the map manifests itself in how he or she actually goes about getting from A to B.
The nub of the matter: focal and subsidiary awareness
Let me get to the heart of the matter through another example that is especially relevant as I sit at my desk writing these words.
I ask the following question:
What is it that enables me to write these lines using my knowledge of the English language, papers on knowledge management and a host of other things that I’m not even aware of?
I’ll begin my answer by quoting yet again from Tsoukas’ paper,
For Polanyi the starting point towards answering this question is to acknowledge that “the aim of a skilful performance is achieved by the observance of a set of rules which are not known as such to the person following them.” …Interestingly, such ignorance is hardly detrimental to [the] effective carrying out of [the] task…
Any particular elements of the situation which may help the purpose of a mental effort are selected insofar as they contribute to the performance at hand, without the performer knowing them as they would appear in themselves. The particulars are subsidiarily known insofar as they contribute to the action performed. As Polanyi remarks, ‘this is the usual process of unconscious trial and error by which we feel our way to success and may continue to improve on our success without specifiably knowing how we do it.’
Polanyi noted that there are two distinct kinds of awareness that play a role in any (knowledge-based) action. The first one is conscious awareness of what one is doing (Polanyi called this focal awareness). The second is subsidiary awareness: the things that one is not consciously aware of but nevertheless have a bearing on the action.
Back to my example, as I write these words I’m consciously aware of the words appearing on my screen as I type whereas I’m subsidarily aware of a host of other things I cannot fully enumerate: my thoughts, composition skills, vocabulary and all the other things that have a bearing on my writing (my typing skills, for example).
The two kinds of awareness, focal and subsidiary, are mutually exclusive: the instant I shift my awareness from the words appearing on my screen, I lose flow and the act of writing is interrupted. Yet, both kinds of awareness are necessary for the act of writing. Moreover, since my awareness of the subsidiary elements of writing is not conscious, I cannot describe them. The minute I shift attention to them, the nature of my awareness of them changes – they become things in their own right instead of elements that have a bearing on my writing.
In brief, the knowledge-based act of writing is composed of both conscious and subsidiary elements in an inseparable way. I can no more describe all the knowledge involved in the act than I can the full glory of a beautiful sunset.
From the above it appears that the central objective of knowledge management is essentially unattainable because all knowledge has tacit elements that cannot be “converted” or codified explicitly. We can no more capture or convert knowledge than we can “know how others know.” Sure, one can get people to document what they do, or even capture their words and actions on media. However this does not amount to knowing what they know. In his paper, Tsoukas writes about the ineffability of tacit knowledge. However, as I have argued, all knowledge is ineffably tacit. I hazard that this may, at least in part, be the reason why KM initiatives fall short of their objectives.
Acknowledgement and further reading
Thanks to Paul Culmsee for getting me reading and thinking about this stuff again! Some of the issues that I have discussed above are touched upon in the book I have written with Paul.
Finally, for those who are interested, here are some of my earlier pieces on tacit knowledge:
1 As far as I’m aware, Nonaka and Takeuchi’s model mentioned in this article is still the gold standard in KM. In recent years, there have been a number of criticisms of the model (see this paper by Gourlay, or especially this one by Powell). Nonaka and von Krogh attempt to rebut some of the criticisms in this paper. I will leave it to interested readers to make up their own minds as to how convincing their rebuttal is.