Reason and Interdisciplinarity in Knowledge Management

Abstract

We take a tour of what a valid and useful knowledge management would look like. To do so we distinguish between the two major types of organizational knowledge – human and technological – and show how rationality is to be differently understood in these two types. Human or personal reason is distinguished by a nonseparability of its concepts owing to the intuitive manner in which they are unified in the act of understanding, so that knowledge is a risky matter of interpretation and judgment. Technology instead requires a separability of concepts essential for reducing them to measurement, where they must stand visibly apart. We go on to show the paradoxical but essential need of these two knowledge types for one another, shown in the duality between unobservable and observable concepts necessary for judging an organization’s value. We find that fact and counterfact, actual and possible, observable and unobservable are brought together in organizational knowledge.

1. Introduction

By one account, knowledge management is the assessment and control of the data and know-how of an organization. Especially though, it is a matter of forming and using the data and knowledge, as only then can they inform both strategy and the processes of carrying it out.

Abstract definitions like this unfortunately only get us started. Few if any organizational improvements will be forthcoming from stated principles and purposes, explaining why our emphasis is usually on case studies, which cut through the generalities and offer a sense or feeling of what might work

Some insight however is to be had from asking what business knowledge is itself. What is that we call knowledge? How does knowledge lead to understanding, which must be knowledge’s final goal? How will we know the best way to build the organization we are engaged with and how can these ‘best-ways’ be made accessible? To address these questions I step behind the important administrative and logistical questions to look in the opposite direction: at the sources of the ideas and concepts by which knowledge brings the understanding that effective management requires. I begin with rationality itself, since in the absence of reason there would be no determinate solution to any organizational problem whatever, only a procession of ad hoc rules justifying any idea we came up with. We can envisage three cognitive steps in knowledge management, a reflection on: (i) rationality and optimality themselves, (ii) the appropriate optimization and performance procedures, and (iii) their logistics and administration. Of these I’ll concentrate on the first while keeping the second and third in mind.

2. Reason: fact and counterfact

Our interest is in knowledge management itself. But what is knowledge? This is a metaphysical as well as epistemological question, bearing on the structure of reality as well as the structure of how we know reality. So it’s worthwhile reflecting on both these questions. Something hangs paradoxical over the topic of knowledge management, at least when the knowledge is technical. For the question of managing something is simultaneous with, co-determined by, the question of how we value it. And technical knowledge is at a disadvantage when raising or answering questions of value.

2.1 Technology does not assess value, even its own

The difficulty is that value judgments can never be a matter of measurement; and it is with measurement that technology principally occupies itself. Reflections on measures and so on value are a matter of judgment. And judgment is a form of reason that transcends – though at the same time includes – any observability or measurability. Some aspects of the task before us therefore, insofar as we are forming judgments about technology management, will not themselves be technological. We must be economists, legal scholars, psychologists, or sociologists as well, and in some respect philosophers. Each in its way is a concern for the possibility of knowledge. As Martin Heidegger [1] put it, philosophy has for the past several centuries consisted mostly of “reflection on knowledge itself and its possibility.” (297) And because, one way or another, the work of the philosopher is to reflect on essences, Heidegger paradoxically cautions us that “the essence of technology is by no means anything technological.” [2].

A prime question in modern knowledge management, both in theory and practice, is the issue of the boundaries between knowledge fields. It is not just a matter of where a boundary lies but more fundamentally in what any boundary would consist. Can, should, and if so how and when might two knowledge disciplines be drawn together, or at least brought into conversation? These are relatively easy questions to answer in fields sharing a common epistemological attitude, where the standards for ‘how can I claim that such-and-such is true’ are the same. It is more problematic when their epistemologies or meanings differ, and even more so when their metaphysics do.

A crucial issue and theme of this essay will be the significance to knowledge management of cognitive nonseparability, which arises when one concept, assumption, or finding cannot be made sense of without considering others. Nonseparability is often indicated in multivariate equations when some interaction terms cannot be factored out of the remaining ones. As we will see, separability is a signal property of technological inquiry, and nonseparability a property found in the specifically human disciplines like economics. Indeed, separability virtually defines technical know-how, as nonseparability does human know-how. Yet the two fields cannot do without each other.

2.2 Technology is judgment-challenged, economics is observation-challenged

To the extent human understanding involves nonseparable concepts, rationality is something we can only dwell on rather than measure, since to measure is to stand outside of, be separate from, the object measured. And as psychologist J. H. Van den Berg [3] asked, “who could draw the line between the ‘objective’ and the ‘human’ world?” (99).

The object is; this means that the object in its abstraction is immediately and continually transcended into a whole of (human) interpretations, a whole which, eventually, always encompasses the entire world.” (98).

Those entire worlds are what confront us, or invite us, in every choice we make. It is what Van Dusen [4] calls being-in-the-world, (37) or as Medard Boss [5] puts it, “a world-unfolding and world-opening being.” (85) A person is not then an ‘individual’ at all if by that is meant dissociable for inspection, which, Boss says, could “only be maintained on the basis of abstract intellectual constructions.” (88).

For economics too, it is this same inner intuition, your sense of yourself-in-a-world, that is the basis of your empirical judgments. This is so not only in economics. An intuition or sense, originating in an inner subjectivity and awareness of one’s self in a given situation, is the center of any rationality. This claim can neither be analytically proved nor empirically depicted. The reverse is so: it is the awareness of yourself and situation, projected onto your possible future, that offers the understandability of what you are reasoning and observing. That is nonseparability in a nutshell.

There is however a literal, empirical, and so entirely separable side to organizational thought – technology. As such, technology brings an important qualifier and co-interpreter of economic reason. Every human thought and decision is accompanied with and influenced by, for instance, emotive affects. Empirical, technical psychology is the study of these affects, as chemistry is of the environment and bioengineering is of gene splicing. Together they constitute part of the setting in which organizational analysis is conducted, the foundations of the supplementary theories necessary for business inference.

Economic models – directed to persons in the immediacy of their sense of a situation – therefore must buttress themselves with controlling factors from these technical disciplines. Immanuel Kant [6], who first clarified this problem in terms a modern ear can appreciate, wrote: “The sensible faculty of intuition is really only a receptivity for being affected in a certain way with representations.” (A494, B522) One can go on and form judgments from these representations. But judgments are so closely connected to the initiating intuition that the reverse is also true: Kant says, “the same function that gives unity to the different representations in a judgment also gives unity to the mere synthesis of different representations in an intuition … [and its corresponding] analytical unity.” (153) The prime implication of these words, and of the individuality or singularity of the person who forms the judgment, is that vague abstractions are finally to be ruled out.

It is you who judge. The ‘I think’, says Kant, “and which is the foundation of one’s conceptualizing, must be able to accompany all my representations.” Thus, “intuition has a necessary relation to the ‘I think’ in the same subject in which this manifold is to be encountered.” (B131–132) (163) Sense and concept go hand in hand and it is in a person that they do. These are nonseparable and why technology alone is never finally adequate to knowledge management, just as abstract ideas and groupings aren’t either. This is not to say that empirics and generalizations cannot be stressed in particular studies. They will be and must be.

2.3 Economic reason unites observables with unobservables

All this was well understood by the late 19th century expositors of modern economics. Leon Walras [7] for instance expressed nonseparability in the economic notion of rarete, a sense of a thing’s scarcity, which remains the center of microeconomic theory. Scarcity combines, nonseparably, what we might otherwise think to be the most separable of all: the desirability of a good versus its availability. These two are never separable in a human decision, influencing one another instead. Each therefore has immediate implications for the other, so you can observe neither desirability nor availability on its own. There are no data on them as such because neither is an ‘it’ that can be isolated. Yet the greater your sense of scarcity the more you are willing to pay for it:

$$Scarcity\Rightarrow Price$$E1

By way of the marketplace on the other hand, prices are observable and measurable. Consequently the quantities of this good are observable and measurable too, for how could a per-unit price be observed if the quantities received for it were not? We also then have the relation:

$$Quantity\Rightarrow Scarcity$$E2

insofar as the greater the quantity on the market, the less its scarcity. Finally we combine the two to form:

$$Quantity\Rightarrow Scarcity\Rightarrow Price$$E3

and simultaneously

$$Price\Rightarrow Scarcity\Rightarrow Quantity.$$E4

Now note something fundamental in Walras’ reasoning: what is unobservable here, the scarcity, is what is giving meaning to what is observable – the prices and quantities. In return it is these two observables that point to, and so help outline and reinforce, the sense of the unobservable: the scarcity. They are essential to each other. At the heart of both is what we would call an openness or receptivity to the decision maker’s situation or ‘world’, the future into which she is always stepping.

The synthetic a priori reasoning Kant helped us recognize in ourselves is the form of reasoning we have outlined here and that Kant calls judgment. In it, our immediate sensible intuition of a manifold is formally (pre-empirically) combined into a single representation so as to render it conceptual and therefore articulable and interpretable. Such judgments are never dogmatic. They are a priori only, so must now be synthesized or tested in the presence of observable, empirical data. Scarcity – part of Kant’s category of Limitation or finitude and in turn an element of what he calls Quality – is one such purely formal concept that provides the foundation of empirical work (B106). Because entirely formal, it is unobservable. But as representing a manifold of possibility, it is that by which we grasp and understand the prices and quantities that we observe and measure.

It is in this spirit that James Buchanan [8] shows in the theory of cost a tentative distinction between ‘subjectivity’, which is a matter of sense, and ‘objectivity’ a matter of concept. In the former, and what Buchanan calls the logic of choice, (45) “cost cannot be measured by someone other than the decision maker because there is no way that subjective experience can be directly observed.” “Cost [therefore] must be reckoned in a utility dimension.” (41) In what Buchanan calls the logic of prediction on the other hand, (40) where objectivity and so measurability are essential, we have “the familiar textbook diagrams, the objectively identifiable magnitude that is minimized.” And as we saw in Walras, “this market value is reflected in the market prices for resource units.”(40) ¹ The unobservable then is being joined to the observable.

The highly specific nature of this conjunction of heterogeneous elements calls for metaphorical language. Buchanan says, “properly interpreted, [cost] theory’s claim is limited to making predictions on the ‘as if’ assumption that men do so behave in some average or representative sense.” (38) Metaphor is the only way, he concludes, that the objective and subjective elements of a theory can be integrated. In cost theory, the objective is only ‘standing in for’ the subjective. The business analyst is only ‘standing in for’ the customer, the stockholder, the banker, the CEO, or the Board who make the final decisions.²

Buchanan and I are employing synthetic a priori reasoning here ourselves, as opposed to the merely analytical or merely synthetical (empirical) one. As Lanier Anderson [10] writes in regard to Kantian epistemology in general, an implication is that,

the distinction between analytic and synthetic judgments is fixed independently from our strategies for forming concepts…. (32).

The distinction between the analytic and synthetic, in how each stands toward or interrelates with the other, is already there in the structure of human reason so cannot be rearranged to suit the reasonings of an arbitrarily identified discipline or school. As Patricia Kitcher [11] discusses also from a Kantian standpoint, we cannot choose to take the analytical side over the empirical side, say, of some question simply by a nudge from the other. We cannot derive an analytical theory from a synthetical finding, or a synthetic theory from an analytical finding. They come together instead in every personal judgment.

2.4 The nature of counterfactual (interpretive) reason

Frank Knight [12] had, in more an economist’s or statistician’s than a philosopher’s language, said as much in connection with the phenomenon of risk. “We have insisted that there is a fundamental difference between ‘a priori’ probability on the one hand and statistical” probability on the other. In the former the chances of something happening “can be computed on general principles, while in the latter they can be determined only empirically [synthetically].” (130) ³ This distinction of course bears on the ability to construct, and on the purpose for constructing, a counterfactual and so of the model that generates it.

Knight writes, “we must infer” by way of a model “what the future situation would have been without our interference, and what change will be wrought in it by our action.” (118) Keep in mind we aren’t speaking of just formal modeling here. Every time you stop and think about a specific organizational situation, you are constructing a model of it, where possibilities predominate over actualities. In sum:

To model something is to generate counterfactuals (possibilities).

To be a counterfactual is to be estimated in a model.

J-C Spender [16] adopts this same view in his assessment of what he calls organizational capital. In an extensive and careful critique of proposals to depict and measure the features of an organization – of say a capital asset – Spender concludes that none stand up to inspection. Indeed he remains a skeptic of any claim to capture these phenomena in mere concepts. Our references to reality must then be indirect, by way of pointing. There is no ‘thing’ out there that is ‘an’ asset, as it is too entwined with neighboring ones. Spender says, of human assets in particular: “From the theory side, one difficulty is the idea that either human or social capital can be conceived, measured or theorized independently of the other.” (8–9) For, “employees are constantly educating each other and increasing each other’s human capital in ways that make it difficult to distinguish the consumption” side of knowledge from the production side of it (6).

The underlying message is that, as in other human fields, business and economic concepts are for the most part nonseparable from one another. Pervasive nonseparability is acknowledged in ordinary qualifiers like ‘sort of’, ‘semi’, and ‘partly’. “Elements of the organization’s knowledge and skills are,” Spender writes, “semi-public goods as far as that particular community is concerned.” (9) Insofar as public, they cannot be reduced to any single ‘I think’, but achieve their public character when one person’s ‘I think’ appeals to another person’s ‘I think’.

Efforts to escape or obscure such cognitive inseparability are frequent in the knowledge management literature. But finance, marketing, and employee relations are never really separate in the way that an assessment of one could simply be added to the assessments of the others. Considerations of each will require re-referencing every other in a step-by-step construction, as Westphal [17] demonstrates. Plans to reorganize a product assembly line, say, will have consequences for a recent bank loan agreement, in turn affecting a proposed new customer product, and around again to the assembly line problem. Joint, nonseparable phenomena of such sort are finally what matter in business, government, and economic life.⁴

We will put this even more succinctly. Synthetic a priori reason, in which a person’s intuitions of a certain setting organize themselves under articulable categories, is nonseparable from intuitions of still other settings – unified as they are by the very nature of intuition. And because all such cognition is a matter of possibility, of a ‘would be’, it is a matter of a counterfactual, a ‘might be’. In short, counterfactual, unobservable-but-estimable, and nonseparable imply one another.

3. Interdisciplinarity: economics and technology

In the above light therefore we ask more generally into the possibility of a valid interdisciplinarity, in which the necessarily mechanical modality of technology might, together with the a-mechanical reasoning of human personhood, help understand human ends and choices.

3.1 Measurement can be unified with concept

Max Weber [15] draws an incisive distinction here:

The term ‘technology’ applied to an action refers to the totality of means employed as opposed to the meaning or end at which the action is, in the last analysis, oriented. (160) The presence of a ‘technical question’ always means that there is some doubt over the choice of the most efficient means to an end. The only considerations relevant are those bearing on the achievement of a particular end …. (161).

Technology for Weber is pure measurable efficiency, the least costly way, in measurable empirical terms, of attaining an end. Not so in economics:

There is, of course, an analogous principle governing economic action. But…it has a different meaning (161). Account must be taken of the quality, the certainty, and the permanence of the result….The definition of economic action must be as general as possible and must bring out the fact that all ‘economic’ processes and objects are characterized as such entirely by the meaning they have for human action in such roles as ends, means, obstacles, and by-products. (158–159).

Economics’ distinctive role is to choose the ends, though the means and their obstacles and byproducts are drawn into them. Ends are chosen for their quality and reliability, characterized Weber says “entirely by the meaning they have” and where prudence or judgment play a role.

What however will comprise quality and reliability? We have no sense in Weber that technology itself will answer this question in any way but an “efficient means” to do something we want done. Furthermore, ‘want done’ is never fully specifiable, extending itself in an indeterminate variety of directions, depending as it does on future contingencies. Technology then demands strict observability, declining to assert ultimate judgment but an element of the judgment’s empirical foundation. For since it is not in a position to know a specific end for a specific person, technology is not in a position to project her future either. Only the organization or business is in the position necessary to appreciate a decision’s ‘by-products’ (opportunity costs) and in that way its meaning. Only with a human view therefore – together with the accompanying data – can one point to the future, to the locus of judgment and purpose.⁵

Let us look more then at what this agency is that offers itself as a valuable auxiliary, that promises to supply us with the most efficient way of achieving whatever ends we have. Importantly, technology is not anything in-itself that you can hold up and literally look at. It is not your cellphone itself. It is a particular interpretation of it: what will help you achieve what you wish. Viewed this way, a phone is thoroughly observable in the literal way that ‘observable’ implies – empirical and measurable.

But it could be interpreted differently too. It could be among the organization’s ends. There it is subject to, combined with, a whole manifold of other ends whose possibilities concatenate without clear limit, one building on another and so necessarily introducing uncertainty. Consider for instance the case of sourcing a newly redesigned cellphone part. It could easily be, compared to others: (a) more expensive, (b) more attractive, (c) more suited to a current advertising scheme, (d) easier to install, (e) more breakable, and (f) not yet fully tested. The data science staff, the engineers, the buyers, the stockers and inventory personnel, the accountants, are now, as technologists, over their heads. You’ve just entered the domain of economic reason if not the legal and sociological. The distinction between ends and means, and so between economics and technology, is itself then a matter of interpretation. And in the dimension of interpretation you are in the dimension of economics and other human, transcendental thought – namely transcending, and therefore judging, what can be literally observed. So if you regard yourself a technologist, keep in mind you spend much of your day a law theorist, economist, and psychologist as well – even if merely in crossing the street – where you are judging the net value of yourself and your organization as an end as well as a means.

As we have made clear and as Kant [6], Heidegger [20], and others show, the distinction between ends and means is co-relative to the distinction between a person and a thing. This distinction in turn is metaphysical because it presumes person and thing to be irremediably disjoint, that a person can never be a thing and vice versa. What it is to be a person then will always be correlative to the question of what it is to be rational, and because rational, conscious of uncertainty. To this uncertainty, technology has of itself no access. Technology is useful for addressing uncertainty but not in itself and as such. That only a person can do, as a person.

3.2 Fact and counterfact are synthetic a priori too: Buyer demand

We can then see why, in declining to put themselves ‘above’ their subjects, the human disciplines must somehow bring together what is observable (e.g. prices, quantities, technology) with what ultimately is unobservable (preferences). Only then can what is specifically human reveal itself. A central way economics does so is in the Marshallian [21] and Hicksian [22] demand functions, the former relating the quantity of a good desired to the money income m of the person desiring it, and the latter to her preferences or utility u instead. Both demands respond also to observable prices p. But while the income in the Marshallian is empirical to both buyer and analyst, the utility in the Hicksian is empirical to neither of them. u is not observable, even to the buyer.

Let us represent these two explicitly, the Marshallian quantity demanded as y (p, m) and the Hicksian as h (p, u).⁶ Now note, in addition to these demanded quantities, that the buyer in the Hicksian case is equivalently wishing, at the given prices p, to be minimizing the expenditures needed to achieve that given u. We will call this minimized expenditure e(p, u). Substituting it for the money expenditure m in the Marshallian demand y(p, m) gives us y[p, e(p, u)] = h(p, u), namely the Hicksian again. That is, because people want to minimize the cost of realizing their yet-unachieved dreams and preferences, we should depict their demand both in the empirical Marshallian dimension and in the nonempirical Hicksian dimension transcending the empirical. Each of these two demands is a check on the other.

Indeed they co-determine each other, since substituting e(p, u) for m – and vice versa – will allow us to move back and forth between the empirical y(p, m) on the one side and the nonempirical h(p, u) on the other. The unobservable is being ‘revealed’ in the empirical, while the empirical is being critiqued and judged by the yet unobservable cost e(p, u) of realizing one’s preferences. Because however we cannot observe u or therefore e(p, u) or h (p, u), the Hicksian ‘pushes’ us to include in the Marshallian the observable factors z that may be correlated with the unobservable but judgeable u. That is, the h (p, u) judgment is what lets us reflect on and specify the y(p, m) more fully, which then, upon estimation, provides the empirical results. Let us refer to this fuller Marshallian demand specification as y(p, m, z).

Despite that such duality or mutual correspondence wasn’t fully clarified until the mid-20th century (by John Hicks [22], in fact), it was implicitly being assumed by the late 19th. For the unobservable as well as observable dimensions of a decision are part of our daily experience. A price for instance is a weight one bears, and so, as a literal heaviness, is fully empirical. And income is countable money multiplied by its countable purchasing power, so equally empirical. Utility in contrast is a judgment of product usefulness or value and hence subject to the conditions of judgment – one’s grasp of a situation and its possible conceptualizations. As possibility only, it is nonempirical and nonobservable while at the same time offering the insight needed to specify the fuller Marshallian demand y(p, m, z).⁷ Here then, in the specification of z and thus of the final demand y(p, m, z) is the opening for psychology, sociology, and political science, along with the technology and engineering needed – by reflecting on e(p, u) – to understand y(p, m, z).

Note that business, law, and economics in particular must now judge the stability and veracity of these other disciplines it is calling in to buttress themselves. Historically the first to be called in was physics, in the form of equation structures flexible enough to distinguish between an identity on the one hand and an equilibration (an insight, an effort, a risk) on the other. The second were the engineers and biologists to detail the technical connections between inputs and outputs, in particular the input combinations x able to produce given outputs q^o under a given technology and from which the cost-minimizing ones are to be selected.

3.3 The hybrid (dual) character of supply

Now that we have settled on consumer demand, consider the hybrid, techno-economic task of the supply side: isolating the input volumes – we’ll call them x – that not only can produce the given q^o but in the least costly way. This is achieved by solving the producer’s cost minimization problem min_xrx(q^o) = min_x [r₁x₁(q^o) + r₂x₂(q^o) + ….,] yielding cost function C(q^o, r), namely the least-cost way of producing quantity q^o at given input prices r.⁸ As you can see, this specifies the desired input volumes (conditional input demands) x(q^o, r) – being the input purchases at alternative prices that minimize the total cost of producing the given output q^o.

Note what we have done here on the producer side, as in the Marshallian and Hicksian on the consumer, is to estimate counterfactuals, consumption and production possibilities under various possible price and other conditions. Economics is more concerned and involved with counterfacts than with ‘facts’.

For all this, technical knowledge has been, though not sufficient, a necessary element. Technical contributions to economic thought and modeling include:

• By Discipline Area: science, engineering, statistics, depth psychology, empirical psychology, and empirical sociology.

• By Application Area: environmental, medical, manufacturing, construction, managerial, and financial.

Importantly, the technical possibilities these fields identify and generate can in turn, if bounded and finite, be recovered from the costs the economic decision maker incurs in their presence. That is to say, we have another duality. Technologies can be reverse-engineered: re-discovered from economic cost data alone, though only if the decision maker had been trying to minimize cost, as only then will the techno-economist’s estimate of the minimum cost, and thus of the technology, be unbiased. It seems no other joint discipline (econ and tech) can accomplish this task – of returning, as digested and absorbed, the very information it had borrowed from another field. And because technology is what is being borrowed-from and interpreted here, and because cost minimization is economics’ mode of borrowing and interpreting, we can say that interpretation is being integrated with the interpreted.

This integration or conjoining of the impersonal and technical with the personal and economic is yet another instance of synthetic a priori reason, a formal schema uniting the decision maker’s inner sense of a situation with his conceptualizations and articulations of it, that Hume [25] first glimpsed and Kant [6] later confirmed. Here then is the nonseparability of what is subjective and what is objective, the nonseparability where real persons are making real decisions.

3.4 Concluding comment: The plight of the counterfactuals

The initial or primal problem we have described here – in which technology is drawn upon to help minimize cost – is, as we have said, a matter of an observable pointing to a nonobservable: technology being the observable and empirical, and economic preferences being the nonobservable and nonempirical. Can you see that the dual problem then – reversing this direction by using the cost-minimizing input combinations to rediscover the technology that had been used – is by this token a matter of the unobservable judging the observable? As we have said, economics and technology are in deep need of one another.

If so, it is an interesting thought experiment to reverse these two poles. It would be to cast economics as the auxiliary, the external information-provider, and technology the interpreter of the information. It would be, that is, for economics (or psychology or sociology) to provide their services to technology. Economic findings then would have to be of the same order as the technical, namely technological themselves. The first thing to see about such a reversal is that the possibility of contextual, personal readings of economic language would be given up. Broad generalizations, emptied of the discrete, intuitive meanings that particular situations provide and that microeconomics is engaged in modeling, then replace them. Persons, the dimension of interpretation, judgment, and decision, are reduced to things.

What might, in an historical sense, explain such depersonalization? Two possibilities its seems to me. First, science and technology afford the unending stream of new curiosities that the human disciplines lack. The plethora of data becoming available in our era is, however, the second reason. Econometric structure, the distinction for instance between endogenous and exogenous variables, initially gives way to single-equation and other reduced-forms, where more abstract visions are allowed. Statistical structure then stands in for theoretical structure. Distinctions between cause and effect devolve into correlations. Generalizations, cut loose from a sense of particular circumstances, get freer rein but at the cost of credibility and meaning.

Yet the management of knowledge, especially of technical knowledge can never finally be technological. If it were, no one would finally be interested as it would have no human implications or meaning. Knowledge management therefore calls for symbolic, ‘as-if’ expression just as economics does. The way technical knowledge is interpreted is finally, as Heidegger [2] was saying, a human question. And to be human is to have a zone of privacy even from oneself. Not even the decision maker can ultimately peer inside and see how her decisions are made, so knowledge managers’ depictions of decision processes are metaphorical and symbolical only. This is not to say metaphor is not effective. Quite the contrary, it is finally the most effective of all. Efforts nevertheless are underway – in artificial intelligence – to eliminate metaphorical, ‘as-if’ expression. That, I surmise, will be AI’s chief problem.

Nomenclature

Epistemology (Methodology)	The study of the methods we use to know something and how the methods can be justified
Counterfactual	A model estimate of what would have happened had conditions been otherwise. In other words, the level an endogenous variable would have taken had the exogenous been at levels different from those at that same time and place
Empirical	Observable with the five senses
Endogenous Variable	A model variable whose value depends on others in the model
Exogenous Variable	A model variable whose value depends on no others in the model but affect others instead
Reduced Form	A single-equation simplification of what is presumed, in fuller reality, to be a multi-equation and more complex one
Separability	An independence or dissociability among variables implying each can be understood without reference to the others
Synthetic a Priori	A modeling procedure that involves both, and forms the unity of, conceptualization and empirical observation