Practice & Context.

Prof. Bourbaki, Inventing frames & diffusing overflows, 2015.

 

A logic of learning

One type of learning is 'learning by doing' in the sense familiar to economists as a downward shift of an average cost curve as a function of cumulative uninterrupted production (Yelle, 1979). There is also learning in the sense of (radical) innovation; Schumpeterian 'creative destruction' and 'novel combinations'. The challenge is to explain the latter type of innovation and learning. That is subject to uncertainty rather than risk. And the point of uncertainty is that it yields the problem of induction or, more precisely, the problem of 'abduction', as Peirce (1957) called it. Quite apart from justifying a novelty once it is achieved, in the 'context of justification', how do you go about the development of novelty, in the 'context of discovery'? How do you go from ('abduce from') and existing working modus operandi to a new one that in the future will turn out to be better but which now is not known? Of all the new things you could think of doing, how do you choose the correct one, or even one that is viable, and how do you know whether you know all the options (Holland et al., 1989)? The problem of uncertainty is precisely that you do not. This connects with and important issue in Austrian economics: how is it that an entrepreneur can be right s/he venture into new areas? In Herbert Simon's terms: if uncertainty precludes the 'substantive' rationality of choosing the best from available options, we need a 'procedural' rationality or heuristic, in the form of some modus operandi that is likely to succeed.

Of course, one possibility for a process of adjustment is random trial and error. This would make economic evolution as blind as biological evolution. And, indeed, the lack of rational evaluation by especially small firms injects an element of randomness that contributes to variety. But although rationality is bounded and entrepreneurship entails an element of gambling, people do think, make inferences and limit risk, and they are not always wrong. So how, as a firm, could one go about 'abduction', and maximize chances of survival? Usually, the answer in theories of abduction is to proceed via 'adjacent possibilities': to project an existing practice into a context that is sufficiently similar to have a chance of success, and allows for some prediction of likely results, while it is sufficiently different to yield novel experience and indications for further change with a chance of success. That principle is extended below into a few basic principles of a 'logic of development': principles that maximize the chance of survival in an economic selection environment. 

The first requirement for survival is ongoing production during adaptation: without it we starve even on the road to success. The second requirement is adaptation to novel opportunities and threats. How to do both? How to reconcile continuity and change? How to combine exploitation with exploration? How to go from utilization of existing resources to the development of new ones? The core principle here is that one should not surrender an existing way of doing things before both the motive and the opportunity for a replacement are evident. Before the need arises such a move would be wasteful, and before the opportunity arises it would be impossible. Thus a certain amount of conservatism is rational, but it can easily become excessive and block innovation. There is a trade-off between the need to adapt and the costs involved in terms of uncertainty of whether novelty will be successful, and uncertainty about the organizational repercussions (March, 1991). To make the step to novel practice, one must be prepared to 'unlearn' (Hedberg, 1981): no longer taking established procedures for granted. Thus a necessary (but not always sufficient) condition for innovation generally is that there is perceived need, mostly from external pressure, a threat to continued existence or a shortfall of performance below aspiration levels, as has been the dominant in the literature on organizational learning (see the survey by Cohen and Sproull, 1996).

The following heuristic principles of development and learning are now derived. The first principle of abduction is that one needs generalisation of a successful practice to novel but 'adjacent' contexts, where it is likely to succeed (so that it satisfies the requirements of ongoing production), while it is also likely to run into its limitations, so that we may discover the boundaries of its validity (so that it contributes to the requirement of exploration). Next, as the practice runs into its limitations, it should be adapted to the local context to solve them. This is the principle of differentiation. Again, attempts to adapt contribute to ongoing production as a condition for survival as well as adaptation. Typically, such adaptations are inspired by comparisons with similar, 'adjacent' practices, which in the given context are in some respects more successful, and elements from these more successful practices are imported into the local context at hand. This exchange of elements from different parallel practices, in a given context, is the principle of reciprocation. In language, reciprocation is the operation of metaphor or analogy. As the practice becomes more and more differentiated across contexts, efficiency losses appear owing to lack of standardization, economies of scale foregone, complexity of ad hoc add-ons. Novel elements inserted from outside often do not fit well in the structure of current practice, and for the full utilization of their potential require a more fundamental restructuring of practices in a novel practice. This is the principle of novel combinations or accommodation. This is where the previous preparatory steps lead to novelty. But then, at first, the novelty is indeterminate. Knowledge is partly or wholly tacit: to the extent that novelty works, it cannot yet be fully explained. Much experimentation is needed to find its best form(s), for the novelty to 'come into its own', and to become standardized in a 'dominant design' (Abernathy and Utterback, 1978). This is the principle of consolidation. After that has been achieved, one can move into the next cycle of development, starting with generalization. What feeds exploration while maintaining exploitation is an alternation of variety of practice and variety of context. The resulting cycle of learning is illustrated in Figure 1. The claim is that this procedure is the best answer to the problem of how to maintain continuity (exploitation) while preparing for change (exploration).

The hypothesis now is that these principles of abduction, being conducive to survival, constitute a fundamental 'logic' or heuristic of development which is applicable, with appropriate elaborations and enrichments, at all levels of learning/development/adaptation, and at the level of individual people, organizations, industries and national economies. Thus it should, among other things, serve to specify the relation between equilibrating (Walrasian, Austrian) entrepreneurship and dis-equilibrating Schumpeterian entrepreneurship. Furthermore, this 'logic' should also help to indicate how in processes of development the different levels of people, firms, industries and national or regional economic systems tie into each other. Of course, the hypothesis has to be argued more in detail and then tested extensively. The hypothesis has been inspired by the work of the developmental psychologist Jean Piaget (1970, 1974; Flavell, 1967) on individual cognition. 

Note that, while in a procedural sense the heuristic is optimal, it need not yield a unique or optimal outcome. It allows for path-dependence and suboptimal outcomes, and the path taken depends on context and coincidence. Different economies can develop different structures. 'Logic' is put between quotation marks because it is a heuristic rather than a logic in the sense of indicating a sequence of stages that is logically or epistemologically necessary. It is a heuristic in the sense that it is generally the best answer to the problem of abduction; the best way of exploring while maintaining exploitation. However, stages will overlap: there is generalization during consolidation, differentiation during generalization, exploration of novel combinations during reciprocation. Stages may occasionally even be skipped, and innovation can occur less systematically, more randomly and spontaneously (Cook and Yanow, 1993), when an obvious opportunity presents itself without much exploration. But as a general rule one needs to accumulate failures to build up the need for change, as well as hints at what directions to look in: indications of what changes could be made with some chance of success.

Bart Nooteboom, Innovation, Learning and Industrial Organization, Camb. J. Econ. 23, 1999, pp.131-133.

Exploration & Exploitation.

Procrastimotion, Ringworld, 2012.

A central concern of studies of adaptive processes is the relation between the exploration of new possibilities and the exploitation of old certainties (Schumpeter, 1934; Holland, 1975; Kuran, 1988). Exploration includes things captured by terms such as search, variation, risk taking, experimentation, play, flexibility, discovery, innovation. Exploitation includes such things as refinement, choice, production, efficiency, selection, implementation, execution. Adaptive systems that engage in exploration to the exclusion of exploitation are likely to find that they suffer the costs of experimentation without gaining many of its benefits. They exhibit too many underdeveloped new ideas and too little distinctive competence. Conversely, systems that engage in exploitation to the exclusion of exploration are likely to find themselves trapped in suboptimal equilibria. As a result, maintaining an appropriate balance between exploration and exploitation is a primary factor in system survival and prosperity.

James G. March, Exploration and Exploitation in Organizational Learning, Organ. Sci. 2, 1991, p.71.

Normal Design.

n.a., Paris dans 20 ans, 1967 (via NDLR).

But focus on technology as knowledge has ramifications beyond the science-technology question. Hugh Aitken, for example, makes it basic to the historical method adopted in his book The Continuous Wave. To recount the early technical and institutional history of radio, Aitken regards "history of technology as one branch of intellectual history or the history of ideas." From this approach, he explains the origins of the history of inventions crucial to radio by examining "the flows of information that converged at the point and at the time when the new combinations came into existence." As the work of Aitken and other historians makes clear, however, the ideas we deal with are not disembodied - they are, as Layton points out, the ideas of people (and communities of people). Emphasis on knowledge thus brings history of technology into symbiotic relation, not only with intellectual history and philosophy, but with social history and sociology as well. Such emphasis is critical, in particular, for understanding technological change, a fundamental concern in one way or another for all these disciplines. As remarked by Rachel Laudan, "shifts in the knowledge of the practicioners play a crucial role in technological development." People who aspire to understand such development - economists and policy makers, for instance - might do well to focus accordingly when they delve (in Rosenberg's graphic phrase) "inside the black box" of technology. If these ramifications are valid, as I believe they are, laying out the features of engineering knowledge very much needs doing.

In addressing this task, I will structure the inquiry around the goal of design. For engineers, in contrast to scientists, knowledge is not an end in itself or the central objective of their profession. Rather, it is, as illustrated by the quotation from the British engineer, a means to a utilitarian end - actually, several ends. Engineering can, in fact, be defined in terms of these ends, as in the following quotation from another British engineer, C.F.C. Rogers:

Engineering refers to the practice of organizing the design and construction (and, I would add, operation) of any artifice which transforms the physical world around us to meet some recognized need.

Here I take "organize" to be meant in the sense of "bring into being" or "get together" or "arrange". The first end, "design", has to do with the plans from which the artifice is built, as in the many drawings (or computer displays) of an airplane and its components. "Construction" (which I shall call "production") denotes the process by which these plans are translated into the concrete artifice, as in manufacture of the actual airplane. "Operation" deals with the employment of the artifice in meeting the recognized need, the related example being the maintenance and flight operations of the airplanes of an airline. Definitions of engineering sometimes mention other ends such as "development," and "applications" or "sales"; these can usually be subsumed under one of the foregoing three, which will be sufficient for present purposes.

Of the three, design is frequently taken as central. Layton, in treating technology as knowledge, takes it as such (with minor mention of other ends). He adds in a later paper, however, that recent attempts among engineers to "reestablish design as the central theme of engineering" are "not without ideological overtones." Other scholars contend that rhetorical emphasis on design by engineers is primarily an attempt to gain status, that "engineers have seized on design as a way to liken their activity to that of scientists, to assert that they too are engaged in creative activity." Whatever the truth of the situation, I will restrict my focus here almost entirely to design. To attempt more would extend impractically an already lengthy study. Great numbers of engineers do, in fact, engage in design, and it is there that requirements for much engineering knowledge originate in an immediately technical sense. Though extaengineering needs - economic, military, social, or personal - may set the original problem, for many workaday engineers things come into focus at the level of concrete design. My emphasis on design, however, should not be taken to imply anything derogatory about other areas of engineering. For a complete epistemology of engineering, production and operation will require equal attention. For the time being, however, my concern will be limited to engineering design knowledge.

"Design," of course, denotes both the content of a set of plans (as in "the design for a new airplane") and the process by which those plans are produced. In the latter meaning, it typically involves tentative layout (or layouts) of the arrangement and dimensions of the artifice, checking of the candidate device by mathematical analysis or experimental test to see if it does the required job, and modifications when (as commonly happens at first) it does not. Such procedure usually requires several iterations before finally dimensioned plans can be released for production. Events in the doing are also more complicated than such a brief outline suggests. Numerous difficult trade-offs may be required, calling for decisions on the basis of incomplete or uncertain knowledge. If available knowledge is inadequate, special research may have to be undertaken. The process is a complicated and fascinating one that needs more historical analysis than it has received.

Design is important here, however, mainly as it conditions the knowledge required for its performance. Knowledge itself forms the primary focus; while requirements from design must be kept in mind at all times as determining that knowledge, details of how the process takes place are secondary. I have never attempted to design an airplane in my entire career as a research engineer (though I participated in planning and designing large aeronautical research facilities). The atmosphere  in which I worked, however, and the knowledge I helped produce, were conditioned by the needs of airplane designers who visited our laboratory. My colleagues and I were keenly and continuously aware of the practical purposes we served. The situation in this book is somewhat similar. Though only one of the historical studies deals directly with the design process, the needs of design play a determining role throughout. 

To keep matters manageable, I shall further limit attention to what can be called normal design. In The Origins of the Turbojet Revolution, Edward Constant defined "normal technology" - "what technological communities usually do" - as comprising "the improvement of the accepted tradition or its application under 'new or more stringent conditions.'" Normal design (my extension, not Constant's) is then the design involved in such normal technology. The engineer engaged in such design knows at the outset how the device in question works, what are its customary features, and that, if properly designed along such lines, it has good likelihood of accomplishing the desired task. A designer of a normal aircraft engine prior to the turbojet, for example, took it for granted that the engine should be piston driven by a gasoline-fueled, four-stroke, internal-combustion cycle. The arrangement of cylinders for a high-powered engine would also be taken as given (radial if air-cooled and in linear banks if liquid-cooled). So also would other, less obvious features (e.g., tappet as against, say, sleeve valves). The designer was familiar with engines of this sort and knew they had a long tradition of success. The design problem - often highly demanding within its limits - was one of improvement in the direction of decreased weight and fuel consumption or increased power output or both. Normal design is thus very different from radical design, such as that confronting the initiators of the turbojet revolution described by Constant. The protagonists of that revolution had little to take for granted in the way that designers of normal engines could. In radical design, how the device should be arranged or even how it works are largely unknown. The designer has never seen such a device before and has no presumption of success. The problem is to design something that will function well enough to warrant further development.

Though less conspicuous than radical design, normal design makes up by far the bulk of day-to-day engineering enterprise. The vast design offices at firms like Boeing, General Motors, and Bechtel engage mainly in such activity. In the words of one reader of this material, "For every Kelly Johnson (a highly innovative American airplaine designer who will figure in chapter 3) there are thousands of useful and productive engineers designing from combinations of off-the-shelf technologies that are then tested, adjusted, and refined until they work satisfactorily." In addition, knowledge for normal design is more circumscribed and easier to deal with. Though it may entail novelty and invention in considerable degree, it is not crucially identified with originality in the same way as knowledge for radical design. My restriction to normal design thus related to both substance and expedience - there are sufficient matters of importance to confront at this stage without opening the Pandora's box of technical invention.

I do not mean to suggest that normal and radical design, and the knowledge they require, can be sharply separated; there are obviously middle levels of novelty where the distinction is difficult to make. The difference, nevertheless, is sufficiently real to serve as a basis for analysis. I likewise do not mean to suggest that normal design is routine and deductive and essentially static. Like technology as a whole, it is creative and constructive and changes over time as designers pursue ever more ambitious goals. The changes, however, are incremental instead of essential; normal design is evolutionary rather than revolutionary. As we shall see, even within such limits the kinds of knowledge required are enormously diverse and complex. The activities that produce the knowledge, unlike the activity it is intended to support, are also something for from normal and day-to-day. 

W.G. Vincenti, What Engineers Know and How They Know It, Edition : Reprint. Baltimore: John Hopkins University Press, 1993, 5-9.

Engineering Knowledge.

Tobias Revell, 88.7, 2012 (via hmkv.de).

Engineering knowledge, though pursued at great effort and expense in schools of engineering, receives little attention from scholars in other disciplines. Most such people, when they heed to engineering at all, tend to think of it as applied science. Modern engineers are seen as taking over their knowledge from scientists and, by some occasionally dramatic but probably intellectually uninteresting process, using this knowledge to fashion material artifacts. From this point of view, studying the epistemology of science should automatically subsume the knowledge content of engineering. Engineers know from experience that this view is untrue, and in recent decades historians of technology have produced narrative and analytical evidence in the same direction. Since engineers tend not to be introspective, however, and philosophers and historians (with certain exceptions) have been limited in their technical expertise, the character of engineering knowledge as an epistemological species is only now being examined in detail. This book is a contribution to that effort.

My involvement in the study of engineering knowledge stems in part from a question put to me by my Stanford economics colleague Nathan Rosenberg over lunch in the early 1970s: "What is it you engineers really do?" What engineers do, however, depends on what they know, and my career as a research engineer and teacher has been spent producing and organizing knowledge that scientists for the most part do not address. My attempts to deal with Rosenberg's question led me therefore - without at first realizing just what I was doing - to examine the cognitive dimension of engineering. Given a long-standing interest in history, it was also instinctive for me to approach the problem historically. To my pleasant surprise, I found myself in step with the work being produced by historians of technology.

In the view developed by these historians, technology appears, not as derivative from science, but as an autonomous body of knowledge, identifiably different from the scientific knowledge with which it interacts. The idea of "Technology as Knowledge" - title of an influential paper by Edwin Layton, one of the view's early champions - credits technology with its own "significant component of thought". This form of thought, though different in its specifics, resembles scientific thought in being creative and constructive; it is not simply routine and deductive as assumed in the applied-science model. In this newer view, technology, though it may apply science, is not the same as or entirely applied science.

This view of technology - and hence engineering - as other than science accords with statements sometimes made by engineers, such as the following by a British engineer at the Royal Aeronautical Society in 1922: "Aeroplanes are not designed by science, but by art in spite of some pretence and humbug to the contrary. I do not mean to suggest for one moment that engineering can do without science, on the contrary, it stands on scientific foundations, but there is a big gap between scientific research and the engineering product which has to be bridged by the art of the engineer." The creative, constructive knowledge of the engineer is the knowledge needed to implement that art. Technological knowledge in this view appears enormously richer and more interesting than it does as applied science.

The newer view comes from the work of historians over several decades. The historiographic development has been examined in an extended study by John Staudenmaier and a shorter review by George Wise. Both come to the conclusion, as expressed by Wise, that "treating science and technology as separate spheres of knowledge, both man-made, appears to fit the historical record better than treating science as revealed knowledge and technology as a collection of artifacts once constructed by trial and error but now constructed by applying science." The evidence to be presented here supports this conclusion. The reality of the distinction is emphasized for me by the fact that the school of engineering at my own university, as at all such institutions, finds it necessary to maintain its own library, separate from those of the departments of physics and chemistry. This separation is more than a convenience. Engineers, though they require many of the same books, journals and documents as physicists and chemists, need others not kept in the science libraries. Despite the historical and institutional evidence for its autonomy, however, the features that distinguish technological knowledge have not been laid out in detail.

The view of technology as an autonomous form of knowledge is bound closely with the debate over the relation between technology and science, which has been a long-standing concern of historians of technology. Staudenmaier sees the view of having emerged out of that debate and become a major theme in itself, with the science-technology relation reduced to a subtheme. Wise regards the science-technology relation as still an organizing issue for research, with the view of technology as a special kind of knowledge defining the technological side of the relation. However that may be, viewing technological knowledge as autonomous leaves the relation between technology and science still open to specification. Technological knowledge then takes its place as a component on one side of what can be called an "interactive model" of the relationship. In this model, which has been summarized concisely by Barry Barnes, technology and science are autonomous forms of culture that interact mutually in some complicated and still-to-be-spelled-out fashion. The nature of technological knowledge constrains but does not define the relationship.

Things look very different if the knowledge content of technology is seen as coming entirely from science. Such a view immediately defines the science-technology relation - technology is hierarchically subordinate to science, serving only to deduce the implications of scientific discoveries and give them practical application. This relation is summarized in the discredited statemen that "technology is applied science". Such a hierarchical model leaves nothing basic to be discussed about the nature of the relationship. A model with such rigiditiy is bound to have difficulty fitting complex historical record. 

W.G. Vincenti, What Engineers Know and How They Know It, Edition : Reprint. Baltimore: John Hopkins University Press, 1993, p.3-5. 

Complex Acronym.

Jungle, The Heat, 2014.

'The common misapprehension is that a messy desk is a sign of a hard worker.'

'Get over the idea that your function here is to collect and process as much information as possible.'

'The whole mess and disorder of the desk on the left is, in fact, due to excess information.'

'A mess is information without value.'

'The whole point of cleaning off a desk is to get rid of the information you don't want and keep the information you do want.'

'Who cares which candy wrapper is on top of which paper? Who cares which half-crumpled memo is trapped between two pages of a Revenue Ruling that pertained to a file three days back?'

'Forget the idea that information is good.'

'Only certain information is good.'

'Certain as in some, not as in a hundred percent confirmed.'

'Each file you examine in Rotes will constitute a plethora of information,' the Personnel aide said, stressing the second syllable of plethora in a way that made Sylvanschine's eyelids flutter.

'Your job, in a sense, with each file is to separate the valuable pertinent information from the pointless information.'

'And that requires criteria.'

'A procedure.'

'It's a procedure for processing information.'

'You are all, if you think about it, data processors.'

The next slide on the screen was either a foreign word or a very complex acronym, each letter in bold and also underlined.

'Different groups and teams within groups are given slightly different criteria that help inform what to look for.'

The Personnel aide was thumbing through his laminated outline.

'Actually there's another example of the information thing.'

'I think they've got it.' 

The CTO had a way of turning one foot out perpendicular to its normal direction and tapping it furiously to signal impatience.

'But it's right here under the desk thing.'

'You mean the deck of cards?'

'The checkout line.'

They seemed to believe their mikes were off.

'Christ.'

'Who'd like to hear another example illustrating the idea of collecting information versus processing data?'

Cusk was feeling solid and confident, as he often did after a series of attacks had passed and his nervous system felt depleted and difficult to arouse. He felt that if he'd raised his hand and given an answer that turned out not to be correct it wouldn't have been that big of a deal. 'Whatever,' he thought. The 'whatever' is what he often though when he was feeling jaunty and immune from attack. He had twice actually asked women out when in this cocky, extroverted, hydrotically secure mood, then later failed to show up or call at the appointed time. He actually considered turning around and saying something jaunty and ever so slightly flirtatious to the noisome Belgian swimsuit model - on the upswing, he now wanted people's attention.

At age eight, Sylvanshine had data on his father's liver enzymes and rate of cortical atrophy, but he didn't know what these data meant.

'There you are at the market while your items are being tallied. There's an individual price for each item, obviously. It's often right there on the item, on an adhesive tag, sometimes with the wholesale price also coded in the corner - we can talk about that some other time. At checkout, the cashier enters the price of each grocery, adds them up, appends relevant sales tax - not progressive, this is a current example - and arrives at a total, which you then pay. The point - which has more information, the total amount or the calculation of ten individual items, let's say you had ten items in your cart in the example. The obvious answer is that the set of all the individual prices has much more information than the single number that's the total. It's just that most of the information is irrelevant. If you paid for each item individually, that would be one thing. But you don't. The individual information of the individual price has value only in the context of the total; what the cashier is really doing is discarding information, which in the cashier runs through a procedure in order to arrive at the one piece of information that's valuabe - the total, plus tax.'

'Get rid of the layman's idea that information is good. That the more information the better. The phone book has lots of information, but if you're looking for a phone number, 99.9 percent of that information is just in the way.'

'Information per se is really just a measure of disorder.' Sylvanshine's head popped up at this.

'The point of a procedure is to process and reduce the information in your file to just the information that has value.'

'There's also the matter of using your time most efficiently. You're not going to spend equal time on each file. You want to spend the most time on the files that look promising in terms of yielding the most net revenue.'

'Net revenue is our term for the amount of additional revenue generated by an audit less the cost of the audit.'

'Under the Initiative, examiners are evaluated according to both total net revenue produced and the ratio of total additional revenue produced over total cost of additional audits ordered. Whichever is the least favorable.'

'The ratio is to keep some rube from simply filling out a Memo 20 on every file that hits his Tingle in hopes of jacking up his net.' Cusk considered: An examiner who filed no Memo 20s ever would have a ratio of 0/0 which is infinity. But the net revenue total would, he reflected, also be 0.

'The point is to develop and implement procedures that let you determine as quickly as possible whether a given file merits closer examination -'

'- that closer examination itself involving some type or types of procedures blended with your own creativity and instinct for smelling a rat in the woodwork -'

'- although at the beginning of your service, as you're gaining experience and honing your skills, it will be natural to rely on certain tested procedures -'

'- a lot of these will vary by group or team.'

'Incongruities on the Master Files, for one thing. That's pretty obvious. Disagreement of W-2s plus 1099s with stated income. Disagreement of state return with 1040 -'

'But by how much? Below what floor do you simply let an incongruity go?'

'These are the sorts of matters for your group orientation.'

Sylvanshine now know that two separate pairs of new wigglers were actually, unbeknownst to them, related, one pair through a liaison five generations ago in Utrecht.

David Cusk was now feeling so relaxed and unafraid that he was almost getting drowsy. The two trainers sometimes established a rhythm and concert that was soothing and restful. Cusk's tailbone was a tiny bit numb from settling back and slumping in his seat, resting his elbow casually on the foldout desk, the heat of the little lamp of no more direct concern than news of the weather somewhere else.

David Foster Wallace, The Pale King, Penguin Books, 2012, p.342-345.

Technical Steps.

Professor Bourbaki, IMG_0291, 2013.

Local practices form the place of origin of novelty and new technical knowledge. New technologies emerge as small technical steps in response to local problems, and only later give rise to new technical trajectories. Thus, both new artefacts and knowledge emerge through localised work. But local technical knowledge does not simply flow to other locations, as if it were a discrete entity. Before knowledge can circulate, it has to be made sufficiently context-free. (…) To create generic knowledge that can circulate, dedicated socio-cognitive work is needed to bring about a process of aggregation. “Aggregation” is the process of transforming local knowledge into robust knowledge, which is sufficiently general, abstracted and packaged, so that it is no longer tied to specific contexts. (…) 

Typical aggregation activities include standardisation, model building, writing of handbooks, formulation of best practices. Also codification, a term used by economists to describe the transformation of tacit into codified knowledge, is part of aggregation. (…) 

Codification means recording in a “codebook” and involves model building, language creation and message writing. While codification highlights the “coding” aspect of dynamics, aggregation also emphasises the “de-localisation” aspect. Aggregation refers to a broader socio-cognitive process of which codification is an aspect. 

Aggregation entails the production of a collective good: abstract knowledge that can be used by others. Because of free-rider problems, participation in aggregation processes is not self-evident. Why would you contribute to the build-up of a collective knowledge reservoir and share experiences with others, especially when they are competitors? Without proper arrangements or incentive structures, such collective goods will not be produced optimally, because they can be used by others who have not contributed to its production (Deuten, 2003). An important arrangement is the creation of intermediary actors, for example, professional societies, industry associations, standardisation organisations. Such intermediary actors may be created when actors perceive themselves as part of an emerging community with collective interests. In that case, perceived benefits of producing a collective good may outweigh perceived disadvantages. 

Intermediary actors may perform aggregation activities, because they have special responsibilities and roles. Standardisation organisations, for instance, are responsible for creating and maintaining a collective reservoir of (standardised) technical knowledge (Schmidt and Werle, 1998). Professional societies and industry associations also stimulate and facilitate the production and circulation of technical knowledge. They may create technical standards, articulate problem agendas, and exchange experiences and findings to further the interests of the (emergent) field as a whole. Also firms that travel between local practices may aggregate knowledge. Engineering firms or sector research institutes, for instance, are hired by other firms to perform certain jobs. They can compare experiences in different locations, reflect on differences and draw general conclusions. They can use this aggregated knowledge for other jobs in different locations. Initially, they may aggregate their experiences for intra-organisational purposes only. But they may also be willing to share (parts of) their knowledge reservoirs to enhance their reputation and visibility in relevant forums. 

Aggregation activities by intermediary actors do not revolve around finding technological solutions for local, specific problems, but rather around the creation, maintenance and distribution of generic, abstracted knowledge that can be used throughout a technological field (Rip, 1997; Deuten, 2003). So there is a division of cognitive labour: practical technical work in local practices, and dedicated aggregation activities to transform local experiences into global knowledge. Intermediary actors work at this global level. Intermediary actors are not always present from the start. They are often created as part of the emergence of a new technical community. Also the creation of an infrastructure for circulation and aggregation processes is important. Such an infrastructure consists of forums that enable (and induce) the gathering and interaction of actors, the exchange of experiences and the organisation of collective action. Examples of such forums are conferences, seminars, workshops, technical journals, proceedings, and so on. The creation of these forums tends to be part of community formation processes.

F. Geels and J. J. Deuten, “Local and global dynamics in technological development: a socio-cognitive perspective on knowledge flows and lessons from reinforced concrete,” Sci. Public Policy, vol. 33, no. 4, pp. 265–275, May 2006.

Réseaux & Prépositions.

Eva Leroi, Post City, 2012 (via NDLR).

Le croisement [RES – PRE] est assez particulier puisque c’est lui qui autorise toute l’enquête. Du point de vue des descriptions de type [RES] tous les réseaux se ressemblent (c’est même ce qui permet à l’enquêtrice d’aller partout en s’affranchissant de la notion de DOMAINE) mais, dans ce cas, les PRÉPOSITIONS demeurent totalement invisibles sinon sous la forme d’un léger remords (l’enquêtrice a le sentiment diffus que ses descriptions ratent quelque chose qui semble essentiel aux yeux des informateurs). Inversement, dans une exploration de type [PRE], les réseaux de [RES] ne sont plus qu’un type de trajectoire parmi d’autres et ce sont les modes qui sont devenus incompatibles, bien qu’on puisse comparer deux à deux leurs conditions de félicité mais du seul point de vue de [PRE]. 

Il n’aura pas échapper aux lecteurs un peu sociologues que ce croisement [RES - PRE] pose un problème de “compatibilité logicielle”, comme on dit en informatique, entre la théorie de L’ACTEUR-RÉSEAU et ce que nous venons d’apprendre à noter [PRE]. À l’évidence, pour pouvoir continuer son enquête, l’anthropologue des Modernes doit maintenant faire le deuil de son penchant exclusif pour un argument qui l’avait pourtant libérée de la notion de domaines distincts. 

Cette théorie avait une fonction critique en dissolvant les notions trop étroites d’institution, en permettant de suivre les liaisons entre humains et non-humains et, surtout, en transformant la notion de social et de SOCIÉTÉ en un principe général de libre association, au lieu d’être un ingrédient distinct des autres. Grâce à cette théorie, la société n’est plus faite d’un matériel particulier, le social – en opposition, par exemple, à l’organique, au matériel, à l’économique ou au psychologique – mais d’un mouvement de connexions chaque fois plus étendues et plus surprenantes. 

Et pourtant, nous le comprenons maintenant, cette méthode a conservé certaines des limites de la pensée critique : le vocabulaire qu’elle offre est libérateur mais trop pauvre pour distinguer les valeurs auxquelles les informateurs tiennent mordicus. Ce n’est donc pas tout à fait sans raison qu’on accuse cette théorie de machiavélisme : tout peut s’associer avec tout, sans qu’on sache comment définir ce qui peut réussir et ce qui peut rater. Machine de guerre contre la distinction entre force et raison, elle risquait de succomber à son tour à l’unification de toutes les associations sous le seul règne du nombre de liens établis par ceux qui ont, comme on dit, “réussi”. 

Dans cette nouvelle enquête, le principe de libre association n’offre plus le même métalangage à toutes les situations, mais doit devenir l’une seulement des formes par lesquelles on peut saisir un cours d’action quelconque. Le plus libre, certes, mais pas le plus précis. 

Je peux maintenant récapituler l’objet de cette recherche. En liant les deux modes [RES] et [PRE], l’enquête prétend apprendre à bien parler à ses interlocuteurs de ce qu’ils font – ce par quoi ils passent et de ce qu’ils sont –, ce à quoi ils tiennent. Cette expression de “bien parler” qui fleure bon l’ancienne éloquence comprend plusieurs exigences complémentaires : 

- DÉCRIRE les réseaux sur le mode [RES], avec le danger de choquer les praticiens qui ne sont pas du tout habitués, en modernismes, à parler d’eux-mêmes de cette façon ; 

- VÉRIFIER avec ces mêmes praticiens que tout ce qu’on dit d’eux est bien exactement ce qu’ils savent d’eux-mêmes, mais en pratique seulement ; 

- EXPLORER les raisons du décalage entre ce que révèle la description en termes de réseau et de préposition et son compte rendu par les acteurs ; 

- enfin, c’est là le plus risqué, PROPOSER une autre formulation du lien entre pratique et théorie qui permettrait de mettre fin au décalage et de redessiner les institutions qui pourraient abriter toutes les valeurs auxquelles ils tiennent, sans en écraser aucune au profit d’une autre. 

Le programme est immense mais, du moins est-il clairement défini, d’autant que chacun de ses éléments fait l’objet d’un test spécifique : 

- LE PREMIER est factuel et empirique: avons-nous été fidèles au terrain en ayant les preuves de ce que nous avançons ? 

- LE DEUXIÈME demande une négociation déjà plus compliquée, ce que l’on appelle la restitution à la fin des enquêtes : sommes nous parvenus à nous faire comprendre de ceux que nous avons peut-être choqués, sans abandonner pour autant nos formulations ? 

- LE TROISIÈME est à la fois historique et spéculatif: avons-nous rendu compte des fluctuations historiques entre valeur et réseau ? 

- LE QUATRIÈME suppose des talents d’architecte, d’urbaniste, de designer autant que de diplomate : dans le plan d’habitat ainsi proposé, les futurs habitants se trouvent-ils plus à l’aise qu’avant ? 

Bruno Latour, Enquêtes sur les Modes d’Existence: une Anthropologie des Modernes, Paris: La Découverte, 2012, p.75-77.

Problem Representation.

M. Minard, Carte figurative des pertes successives en hommes de l’Armée Française dans la campagne de Russie 1812-1813, 1869.

For external problem representation, we have provided a simple distinction between sentential representations, in which the data structure is indexed by position in a list, with each element “adjacent” only to the next element in the list, and diagrammatic representations, in which information is indexed in a plane, many elements may share the same location, and each element may be “adjacent” to any number of other elements. While certainly not the complete story on this important representational issue, this simple distinction lets us demonstrate the following reasons why a diagram can be superior to a verbal description for solving problems:

 - Diagrams can group together all information that is used together, thus avoiding large amounts of search for the elements needed to make a problem-solving inference. 

- Diagrams typically use location to group information about a single element, avoiding the need to match symbolic labels.

- Diagrams automatically support a large number of perceptual inferences, which are extremely easy for humans. 

None of these points insure that an arbitrary diagram is worth 10,000 of any set of words. To be useful a diagram must be constructed to take advantage of these features. The possibility of placing several elements at the same or adjacent locations means that information needed for future inference can be grouped together. It does not ensure that a particular diagram does group such information together. Similarly, although every diagram supports some easy perceptual inferences, nothing ensures that these inferences must be useful in the problem-solving process. Failing to use these features is probably part of the reason why some diagrams seem not to help solvers, while others do provide significant help.

J. H. Larkin and H. A. Simon, “Why a Diagram is (Sometimes) Worth Ten Thousand Words,” Cogn. Sci., vol. 11, no. 1, pp. 65–100, Jan. 1987, p.98-99.

Objet Technique.

Gilbert Simondon, Entretien sur la Mécanologie, 1968.

Jean-Yves Chateau a raison de le rappeler dans la longue introduction qui ouvre ce recueil : le souci de l’invention, chez Simondon, n’est pas une manière de dramatiser l’histoire des techniques, c’est une véritable méthode de recherche et d’analyse - mieux, un critère de ce qui est proprement technique, de ce qui fait de la technique "un ordre original de réalité" (p. 14). Aussi l’invention dans le domaine des techniques ne relève-t-elle pas à proprement parler d’une investigation psychologique "au sens habituel du terme", comme le précise Simondon (p. 332). Elle ne se confond pas avec la "créativité" de l’inventeur au travail ; elle ne peut apparaître que rétrospectivement, dans les gestes matérialisés, stabilisés en procédés, dans les objets inventés et les indices matériels de leur élaboration (schémas, prototypes, etc.). On ne lit pas à livre ouvert dans l’esprit des inventeurs : l’esprit est une boîte noire. Qu’il s’agisse du fil à couper le beurre ou de la turbine Guimbal, l’invention doit se lire dans les "traces". La psychologie de l’invention technique suppose donc une forme d’archéologie, et le sujet de l’invention, qu’il soit individuel ou collectif, laborieux ou génial, est toujours un sujet reconstruit à travers ses objets - un sujet technologique plutôt que psychologique. Ainsi la psychologie se confond finalement avec une phénoménologie des objets techniques dont les chemins ne cessent de recroiser ceux de la technologie et de l’histoire des techniques. Mais elle n’est pas séparable non plus d’une ontologie qui interroge le "mode d’existence" de l’objet technique à partir de ce qui fait proprement sa technicité (6). On objectera peut-être ici qu’à moins d’être technicien ou technologue, nous n’avons quasiment jamais affaire à des "objets techniques", mais d’emblée à des ustensiles ou à des machines dont le mode d’être n’est pas séparable de modes d’emploi ou d’usages particuliers. Il y a trois manières de répondre à cela. D’abord, l’objection concède à Simondon le point essentiel, à savoir que les objets techniques n’ont pas l’évidence qu’on leur prête habituellement. À strictement parler, nous ne savons pas ce qu’est un objet technique ; nous ignorons ce qu’il y a de spécifiquement technique dans les objets artificiels dont nous usons le plus souvent sans y penser. L’objet technique est-il d’ailleurs un objet ? Les catégories ordinaires qui nous servent à déployer les modes de l’objectivité ne nous masquent-elles pas l’essentiel ? Ensuite, au niveau fondamental où Simondon prend les choses, le point de vue de la fonctionnalité ou de l’utilité nous détourne de ce qui est proprement technique. L’objet technique n’est d’ailleurs pas nécessairement un outil ou un instrument : il peut être un ustensile, ou une machine présentant des degrés de complexité variables. Or un balai, un aspirateur, peuvent bien servir tous deux à ramasser de la poussière, cet usage commun ne les rapproche pas davantage que le fait de voisiner dans un placard. Enfin, si l’outil et l’instrument remplissent en effet une fonction médiatrice, une fonction de couplage entre un organisme et son milieu, l’essentiel pourtant n’est pas dans ce couplage et les diverses fonctions qu’il remplit : fonction de prolongement (cas de la pince à long bec), fonction de transformation (bras de levier de la pince), fonction d’isolement (pince gainée). Leroi-Gourhan, parmi d’autres, a produit des descriptions et des classifications précises des formes fondamentales de la médiation opérée par l’outil. Cependant, tout reste à faire pour ce qui est d’isoler les critères de la "technicité" et de cerner la nature de l’objet technique. Car "la fonction relationnelle n’est pas la seule : même au niveau le moins élevé, les objets techniques ont une logique interne, une auto-corrélation sans laquelle ils ne pourraient exister" (p. 91).

Elie During, Simondon au Pied du Mur in Critique, 706, 2006.

Convergence Asymétrique.

Mathias Théry, La Vie Après la Mort d'Henrietta Lacks, 2004.

L’idée selon laquelle la recherche n’est plus une activité intellectuelle et solitaire est ancienne. En 1971, Jerry Ravetz décrivait déjà une "industrialisation" du travail scientifique qu’il définissait ainsi : "L’atmosphère sociale devient de plus en plus "industrielle" avec une grande organisation, dans laquelle la main-d’œuvre est affectée à des tâches spécialisées, produit le type de résultats pour lesquels les directeurs ont pu obtenir des contrats avec les agences qui investissent dans une telle production (28)". D’autres auteurs ont également décrit une "convergence asymétrique" entre la recherche publique et la recherche privée (29), caractérisée non seulement par des emprunts réciproques dans les modes de fonctionnement, mais aussi par une prédominance in fine des normes (de rationalisation, de performance, de rentabilité, etc.) du privé. 

La montée en puissance des financements sur projet a accentué certaines formes "d’industrialisation". Au quotidien, la confrontation entre logiques professionnelles et managériales prend notamment forme au travers de l’inflation des tâches administratives de gestion, dont certaines sont au cœur de l’activité de recherche, alors que d’autres relèvent davantage de la routine bureaucratique (30). Les premières englobent notamment le travail de veille et de montage des projets : la présence dans différentes commissions et comités susceptibles de distribuer des fonds, la recherche d’informations concernant les programmes, leurs critères d’évaluation et les chances de les obtenir, la recherche de partenaires et le travail pour constituer des réseaux ou des "consortiums", ainsi que l’ensemble du travail de mise en forme des projets scientifiques destinés à être évalués. Le choix des partenaires s’avère être particulièrement stratégique, non seulement en termes de complémentarité fonctionnelle, mais aussi en termes de visibilité, de reconnaissance et de possibilité d’appropriation des résultats par des partenaires hors du monde académique. 

D’autres pratiques jugées plus périphériques, notamment celles consacrées au suivi plus routinier du projet, induisent une charge de travail qui n’est plus disponible pour d’autres activités jugées comme plus stratégiques, ou faisant partie du "cœur de métier" du chercheur : justifier des dépenses réalisées, mobiliser les partenaires pour organiser les réunions d’avancement ou pour produire les rapports d’avancement des projets, rendre compte des activités réalisées selon les formats préétablis par les agences de financement, etc. Ces exigences d’accountability (31), supposées rendre les organisations transparentes et responsables, pèsent sur le travail des chercheurs titulaires, qui y consacrent une partie croissante de leurs activités quotidiennes.

Hubert Matthieu et Louvel Séverine, Le Financement sur Projet : Quelles Conséquences sur le Travail des Chercheurs ? in Mouvements, 2012/3, n°71, p.22

Notes.

28. J. Ravetz, Scientific knowledge and its social problems, Oxford University Press, Oxford, 1971, p.22.

29. D. L. KLeinman, S. P. Vallas, "Sciences, Capitalism, and the Rise of the "Knowledge Worker" : The Changing Structure of Knowledge Production in the United States", Theory and Society, 30(4), 2001, p.451-492.

30. A. DaHan, et V. Mangematin, "Recherche ou temps perdu ? Vers une intégration des tâches administratives au métier d’enseignant-chercheur", Gérer et Comprendre, Annales des Mines, 102, 2010, p.14-24.