宜昌钢模板租赁:Tacit knowledge: philosophic roots and role in KM

Nonaka and Takeuchi‘s theory borrows from Michael Polanyi‘s theory of tacit knowledge. Most researchers who make use of the concept cite Polanyi or Nonaka and Takeuchi. His theory, therefore, constitutes the starting point for much of the KM literature. A clearer exposition of it should shed light on the advantages and disadvantages of making tacit knowledge a central concept in KM research.

Polanyi developed the concept of tacit knowledge as part of a problematic directed at a philosophic conception of science and scientific theorizing prominent in his time and still influential today. (Polanyi, 1958, 1966, 1969) The theory under attack, logical positivism, held that scientific knowledge was a purely objective matter, being constituted by logically specifiable rules of inference and verification. Scientific hypotheses could be understood as generalizations with specifiable empirical consequences that could be tested using objective methods and devices. Logic, mathematics, objective truth conditions, and a precise vocabulary were the materials of which science was made according to the positivist conception. Polanyi, by contrast, argued that science had a subjective side; that it was not a purely objective affair. His theory of tacit knowledge was the centerpiece of that argument. It was the element in scientific thinking, and in thinking in general, that undermined the claim to pure objectivity.

A brief overview of Polanyi‘s characterization of tacit knowledge is in order. Intuitively, it was expressed by saying that we know more than we can say and it is exemplified by acts of knowledge in which the persons performing the act successfully cannot fully account for their success. For example, a doctor diagnosing a patient by looking at an x-ray cannot say how he or she recognized that the x-ray is abnormal but can just see it based on experience and training. Likewise, when a person recognizes the face of a past acquaintance, he can express his recognition by saying that he knows the person but he cannot describe (fully) how the perceived features of the person‘s face lead to his recognition of him.

These intuitive characterizations are imprecise, however, and do not go very far in conveying the meaning of the concept. Rather, to understand the conception of tacit knowledge, one has to look at the key defining terms Polanyi employed when developing it. The most important terms in his theory are “focal knowledge” and “subsidiary knowledge”. They correspond to the distinction between explicit and implicit knowledge. Focal knowledge is the knowledge that one is directly attending to when one makes a claim to knowledge of some sort. For example, the recognition that an x-ray indicates a disorder or the recognition of a face. The subsidiary knowledge is knowledge that is present in the mind but not attended to directly that leads to the focal knowledge. It is attended to indirectly (in some way) and in relation to the focal knowledge such that it makes the focal knowledge possible. Hence, the defining characteristics of subsidiary knowledge (implicit knowledge) are:

1. that it is active in the mind but not consciously accessed in the moment of knowing; and
2. that it grounds, enables, causes, or somehow brings about the focal knowing.

This example suggests some other concepts used to flesh out the tacit/explicit distinction. These are specifiability, reversibility and “logical gaps”. An act of knowing is specifiable when the steps or processes that lead up to it can be identified. It is reversible when these steps can be formulated as instructions that one could follow to repeat the process. The idea of a logical gap is a break in such a reversible specification that prevents a person from being able to get to the solution to a problem by following a set of steps. In the example above, the tapping that leads to identification exhibits these concepts. How it leads to the identification is unspecified and hence not reversible. There is a logical gap between the initial tapping and its transformation to an enabler of identification.

This logical gap is critical to the theory of tacit knowledge because it is tacit knowledge that plays the role of filling it. What that means is that if we look at the example of tapping, the identification of tactile sensations felt through the hands underdetermines the characterization of tacit knowing. There are, of course, background beliefs and memories in play that allow the sightless person to identify the chair. These beliefs and memories play a similar role to the tactile sensations and hence can be considered a fuller description of the stock of tacit knowledge active in the identification of the chair. What is much more important, however, is that there is an act of mind that integrates the sensations, belief, memories, etc., in such a way as to lead to the identification of the chair. The collection of sensations and other representational states are not by themselves sufficient to make the recognition. Some unspecified operation of the mind unifies them into the extended awareness that crosses the logical gap from one kind of awareness to another. Tacit knowledge is therefore not just subsidiary mental states as described above but whatever it is that unifies them and transforms them so that they fill the logical gap.

One last point about the core theory of tacit knowledge deserves mention. Tacit knowledge, explained in terms of focal and subsidiary awareness, is a structural concept. It describes a relation between different kinds of knowledge and implies unknown principles of operation. The concept is neutral as regards what types of knowledge can stand in the relation. Hence, we can have beliefs, tactile sensations, visual perceptions, etc., playing the role of implicit knowledge. That means that the implicit/explicit distinction does not easily correlate with one kind of knowledge versus another.

KM research tends to treat tacit knowledge as the target of KM practice. Capturing tacit knowledge is seen as the challenge to organizations that want to spread knowledge throughout the organization or spur greater innovation. It is treated as a reserve deposited deep within the ground that needs to be detected and then pumped out. Explicit knowledge, by contrast, is treated as a kind of surface pool that is easier to detect and capture, but which represents only a fraction of the organizational knowledge. Theorists differ on the nature of tacit knowledge. Nonaka and Takeuchi describe it as subjective and mental, as opposed to being objective and external. Hence, to continue the reserve metaphor, they treat the deep buried reserve as having different chemical properties or being in a different physical state. Merely pumping it out will not suffice to make it useful. It needs to be processed and converted into a new form.

Other theorists, for example Nissen, do not require a substantial difference between tacit and explicit knowledge. Whatever it is that is buried is the same as what is in the visible surface pool. Tacit knowledge is like a hidden reserve of water underground, no different from the surface pools, just difficult to get to because of its location and obstacles surrounding it.

Both these conceptions are overly simplistic and may lead to confusion in knowledge management efforts. A common problem they share is that they treat tacit knowledge as something that has value in and of itself. As we saw above, however, the explicit/tacit distinction is structural. It describes a relation between kinds of knowledge. Explicit knowledge depends on tacit knowledge. That is to say, for any explicit knowledge K_e, there is some tacit knowledge, K_t. Further, explicit knowledge is an extension or projection of tacit knowledge to a new level or awareness. Hence, if there is value in identifying tacit knowledge, it is in relation to making explicit knowledge understandable.

Given the structural dimensions of tacit knowledge, the knowledge creation model at the heart of Nonaka and Takeuchi‘s theory needs re-examination. In particular, the conversion process of tacit into explicit knowledge is seen as a process of externalizing individual, subjective knowledge, into publicly sharable knowledge, with some loss of precision along the way. The images in the individual‘s mind are translated into language through the use of metaphor and analogy. While metaphor and analogy are powerful ways to express ideas, they are not the inevitable outcome of an implicit to explicit transition. The structural relation between the implicit and explicit does not require that implicit knowledge consist in images. Implicit knowledge could consist in linguistically expressible beliefs. How the implicit knowledge leads to explicit knowledge in the individual‘s mind is, for Polanyi, the unrecoverable subject element, though that contention is debated in AI and other academic fields.

The more important point, however, is that the tacit/explicit distinction is structural and applies to the thinking processes of the individual. Any explicit knowledge implies tacit knowledge. Nonaka and Takeuchi‘s model of knowledge creation, however, holds that tacit knowledge is converted into explicit knowledge and that the conversion is from the individual to the organization. The individual externalizes implicit knowledge, thus making it available to the organization. However, if the individual has knowledge to contribute to the organization, it must already be explicit. How else would the individual or the organization know that there was some knowledge to be captured in the first place if it were not explicit? Tacit knowledge is not something an individual can work with on its own and by itself it may have no value. It lies behind and enables explicit knowledge.

The above difficulties apply to another part of knowledge creation model: the tacit to tacit transfer of knowledge. If tacit knowledge is intrinsically different from explicit knowledge and needs to be converted, the possibility of tacit to tacit transfers may seem plausible. We can think of this as a knowledge transfer that bypasses explicitness completely and goes directly from one mind to that of another or others. Tacit knowledge flows over a separate channel. This idea is at the heart of much KM writing and involves the concepts of communities of practice, teams, and micro-communities of knowledge. (Von Krogh et al., 2000) But the transfer mechanisms involved are not well characterized in the literature and do not seem to fit the tacit\explicit distinction.

Presumably, tacit knowledge is exchanged when people get together in small groups and talk, exchange ideas, write on white boards, and perform other activities together. The problem is that conversing about a topic is just as much an instance of explicit knowledge as writing out the ideas. If voice capture and translation technologies are used in these settings, the practical difference between speaking and writing disappears. Showing others how to do things should also count as explicit knowledge and it too can be recorded by video or pictures.

This is not to say that tacit knowledge does not play a role in micro-communities. Tacit knowledge plays a role in all individual and group thinking. Rather, the concept of tacit knowledge does not apply to knowledge that is communicated. It applies to the thinking processes of individuals. When groups get together, knowledge is exchanged and new knowledge created. The informal character of these groups raises challenges and opportunities different from formal group activities such as training, carrying out business processes, etc. However, the idea that the difference is spelled out in terms of a tacit to tacit transfer of knowledge as opposed to an explicit to explicit or explicit to implicit transfer is confusing and impossible to model.

The approach of Nissen and others who do not require that tacit knowledge be different from explicit knowledge has its drawbacks as well. The problem also comes from ignoring the structural relationship that defines tacit and explicit knowledge. If tacit knowledge is just knowledge that is in the head, articulating it should be relatively easy, though time consuming. Managers will simply need to allocate time for debriefing, documenting, or decanting. However, tacit knowledge is not always easy to recover because it lies in the background of our thinking and makes that thinking possible. Weakening the sense of tacit knowledge deprives it of its explanatory power as a factor in knowledge transfer that can be appealed to account for success or failure of attempted transfers.

The concept of tacit knowledge is used to represent a number of challenges that companies face in managing knowledge. It is able to do so because its philosophic roots are deep and authors such as Nonaka and Takeuchi have recognized the power of the concept to explain many of the problems organizations face. However, it has come to represent too much and has therefore lost much of its precision and usefulness. Returning it to its philosophic roots, the author believes, can help restore its usefulness and will help in the task of distinguishing different but related problems so that they can be properly addressed. What follows is a brief recap of the tacit/explicit distinction and a discussion of its proper application. The author will lay out a number of points central to the application of the concept in KM theory and practice.

As mentioned above in a number of places, the tacit/explicit distinction is a structural or relational distinction. It does not distinguish between different kinds of mental states (imagistic versus linguistic), but rather it describes the way in which mental states support other mental states and allow them to be acts of knowing. The fundamental idea is that we use ideas to understand or create new ideas, and because our focus must be on the creating and understanding we cannot be aware of all the ideas actively participating in the act of knowing. Sometimes these mental states are experiential, sensual, and emotional. When that is the case, expression in a natural or formal language is difficult if not impossible.

Sometimes, however, these mental states are linguistic and abstract. For example, an engineer working on a technical problem arrives at the answer through a not fully conscious review of a wide range of engineering concepts and techniques, many of which can be documented and were probably learned through reading and reviewing books and articles. That does not make these ideas any less tacit in relation to the explicit knowing. What makes them tacit is their background support and enabling of that act of knowing.

As described above, Polanyi expresses this structural dimension with the terms “focal” and “subsidiary” knowing. The focal knowing is the act of knowing, whether it is a discovery of something new or the application of something learned. We are focused on this knowing and aware of it. For this reason it is described as focal. The subsidiary knowing is that active background knowledge that our mind uses in the act of knowing something. This suggests the following points:

If the last point is correct, a shift in implementation methodology is called for. Much of the research suggests that the proper methodology in a KM initiative is to execute a two-stage knowledge process. In the first stage, an organization inventories all the available explicit knowledge. It then classifies, codes, and stores this knowledge, usually with the help of computer information systems. In the second stage, the organization turns to the more arduous task of identifying all the tacit knowledge residing in the heads of its knowledge workers. Once identified, it attempts to make the knowledge explicit through a conversion process. It can then classify, code, and store this knowledge. The underlying assumption behind this methodology is that tacit knowledge can be pursued by itself, independently of explicit knowledge.

A methodology that respects the structural dependence of explicit and tacit knowledge would look different. On this methodology, an inventory would be performed to determine what explicit knowledge is available and worthy of capture. The results of the inventory would be a list of types or instances of explicit knowledge, (K¹_e,... Kⁿ_e). Next, after each type of explicit knowledge is identified in a list, an analysis would be undertaken to determine what kinds of implicit or tacit knowledge are required to capture and transfer that explicit knowledge. The result would be a set of underlying forms of tacit knowledge that enable or support the explicit knowledge in question, which could be represented as follows (K_e/(K¹_t,…Kⁿ_t)). Having this list, however, would not be enough. In order to facilitate a transfer of K_e, the structural relations between the tacit knowledge set and the explicit knowledge would need to be understood. Once the underlying tacit knowledge is identified and its structure understood, the process of classifying, coding, recording, storing, etc., can take place.

Figures 1 and 2 illustrate the difference between the current implementation model and the new one suggested. Figure 1 depicts the current model.

Under the current model, one goes through a two-step process of identifying tacit knowledge and then converting it to explicit knowledge. Tacit knowledge is treated as structurally independent of explicit knowledge, so after its conversion, it is just added to the inventory of explicit knowledge.

On the new model, we see a structural dependence between explicit and tacit knowledge. Figure 2 depicts this relationship. Here we see a two-step process. In the first step, the explicit knowledge that an organization wishes to transfer and use is identified. In the second step, the tacit knowledge that it depends on, as well as the structure of that dependence, is identified and mapped. The map is used to formulate a plan for codifying and transferring this knowledge.

Using the concepts of specifiability and reversibility, we can more fully characterize the constraints and conditions of knowledge transfer. First, the schema (K_e/(K¹_t,... Kⁿ_t)) presupposes a complete specification of the underlying tacit knowledge relative to some explicit knowledge. That is a theoretical idea. In practice, the specification will always be partial. Time constraints and the nature of the knowledge in question will pose practical limitations on how fully the tacit knowledge can be specified. A major part of the analysis, therefore, will be a determination of the limits of knowledge transfer. If some of the underlying tacit knowledge is not recoverable, limitations will be predicted for any attempted knowledge transfer of the target explicit knowledge. Also, if some tacit knowledge is identifiable, but does not lend itself to linguistic representation, some other form of representation or recreation will need to be identified as part of the transfer strategy. The concept of specification, therefore, serves as a constraint and predictor relative to a knowledge transfer.

The concept of reversibility is also a constraint and predictor. It is closely related to specification but not identical to it. If a full specification is possible, then reversibility falls out from it and ends up differing from specification in its rhetorical or narrative character rather than in substance. However, where full specification is not possible, reversibility may still be achievable. The reason is that the concept of tacit knowledge explains acts of explicit knowledge. That is to say, it explains how we come to know something, whether that is a discovery, an innovation, or a recognition that some item falls under knowledge previously attained. In the context of an organization, it will be critical to determine what particular type of explicit knowledge is the target of transfer.

In the case of an innovation, it is not necessary to recreate the knowledge that lead to the new knowledge. That knowledge may be too extensive and varied to be specifiable. However, once the innovation is discovered, it may be possible to lay out the steps needed to “rediscover” or use the innovation. Students of mathematics, for example, learn proofs for theorems discovered by previous mathematicians. By learning the theorem, they are learning the steps to the discovery. However, the first person to prove the theory did not have those steps. That means that the tacit knowledge he or she had to draw on was different, and more extensive, than the tacit knowledge needed to learn an existing proof. Reversibility, therefore, may be possible, even where a full specification is not.

The above points lead to a revised understanding of the role of tacit knowledge in KM. The tacit/explicit distinction describes a structural relationship between what we know (explicit) and what that knowledge is based on (implicit). Knowledge transfer, therefore, can be best understood as targeting explicit knowledge. However, a constraint on transferring knowledge is the underlying implicit knowledge that supports the target explicit knowledge. Tacit knowledge is an enabling condition of explicit knowledge and of the transfer of knowledge. Organizations attempting to transfer knowledge must therefore identify the relevant tacit knowledge if the transfer is to be successful. The extent to which this is possible will be a predictor of success or failure.

However, since it is easier to learn what is already discovered as opposed to rediscovering the same knowledge, the tacit knowledge required to transfer explicit knowledge will not be as great as that required to discover it in the first place. The demands of reversibility, which correlate directly with transferability, are not as high where diffusion of knowledge is the goal as they are where transfer of innovative capability is attempted. The concept of tacit knowledge helps us understand the challenges involved in both types of transfer and the differences in difficulty between them.