Performance of a complex skill involves, we have remarked, the inte· gration of a number of more elementary units of action. Often, these more elementary units constitute subskills that are optional compo nents of the main skill, selected in response to cues in the perform er’s environment. Thus, the integration required is not just a matter of the relation of the subskills to one another, but also of their rela tion to information arising from the environment. Further, the s ame observations apply to the subskills: they involve integration of still more elementary units, or “subsubskills,” and the integration may again involve relations with the environment as well as within the units. Continuing this descent through the hierarchical structure of the main skill, one comes ultimately to a domain of n eurological and physiological considerations for which the “subskill” terminology is not really appropriate- but reducibility to still more elementary units of action remains possible.
Because skills are such complex, structured entities, and also be cause of the considerations that limit the articulation of the knowl edge applied in a skillful performance, there is inevitably some ambiguity regarding the scope of a skill. This ambiguity has two as pects. There is, first of all, what may be termed operational ambigu ity. It involves predictive uncertainty as to what a particular individ ual who possesses “the skill” can actually accomplish in an attempt to exercise that skill under particular circumstances. The second as pect is the semantic ambiguity of the skill name , the uncertainty regarding the denotation of the term. Operational ambiguity is ob viously one source of semantic ambiguity: to be uncertain a bout whether a particular electrician, functioning as an electrician, will be able to bring about a desired result under particular circumstances is to be a bit vague about what it means to be an electrician. What is more important, semantic ambiguity arises in discussions that ab· stract from the p articular possessor of the skill and the particular cir cumstances of its exercise . Uncertainty about what an electrician is arises in large part from the diversity of electricians and the diversity of tasks and circumstances involved in the exercise of the skills of an electrician.
Both sorts of ambiguity are subject to reduction by deliberate ef fort to that end. By considering the past performances of a particular possessor of the skill, and the characteristics of the particular circum stances in question in relation to those that surrounded the past per formances, it may be possible to sharpen predictions concerning the specific instance. By extending the discussion to subskills, p articular tasks, and quality differentials among possessors of the skill, some of the ambiguity that surrounds the generic skill name can be elimi nated. However, neither of these sorts of clarification is costless, and neither can be totally effective. Both require detailed knowledge of the skill in terms of the mix of subskills involved, the preconditions of effective performance, and so forth. To the extent that this sort of knowledge is tacit, only a person who possesses the skill himself is likely to be in a position to reduce ambiguity by the methods described. To the extent that there are preconditions for effective performance that are simply unknown, or that the tacit knowledge underlying actual performance cannot be brought to bear on the more abstract tasks of assessment and prediction, som. e part of the ambiguity is simply irremediable.
To amplify these points somewhat, consider again the example of the ability to drive a car. This skill is not just the ability to make the vehicle follow a desired course with acceptable accuracy, but also the ability to use a wide range of cues in the environment- other ve hicles, traffic signs and lights, and so on- as the basis for deter mining the details of the course itself. The integration and coordina tion involved in the skilled performance as a whole is not merely of the sort represented in taking a curve smoothly through the coordin ation of pressure on the accelerator and turning of the wheel, but also the relatively automatic use of a large store of information as the basis for interpretive intermediation of sensory input and muscular response. In ordinary discussion about driving, we have little occa sion to attend to the complexity of the skill and the implications of that complexity for the variability of specific driving performances across individuals and across situations. We treat the ability to drive as a dichotomous variable, assuming that the skill is possessed in satisfactory degree or not at all, and regard driver training as a process that transfers individuals from the “unskilled” category to the IIskilled.” This way of talking and thinking about driving skill is typically adequate and we have no need to belabor the complex ities and distinctions of the matter. Occasionally, though, distinc tions are confronted. If a teenage son or daughter is planning a trip with friends, we may concern ourselves with experience levels, atti-tudes toward taking chances, and specific experience with passing on two-lane roads. We may need someone to run an errand and have only a stick- shift car available, and confront the question of whether the assembled ” drivers” include anyone who can shift gears. In such cases, we drop our habitual, implicit homogenization of driving skill and -with the aid of a good many additional words- articulate the distinctions that concern us regarding subskill mixes and so forth.
Sometimes, however, highly relevant distinctions escape con scious consideration or effective articulation. Adverse effects on per formance may arise from causes that do not announce themselves. The ability to control a skid on an icy road will not come in for timely consideration if it is not expected that the roads will be icy. What is not identified cannot be considered, and what is not anticipated cannot be considered in advance. But even fully identifiable and anticipated causes of performance change can resist effective consid eration because of the tacit basis of skill. Consider the American driver who, after the overnight fl ight to London, confronts for the first time the problem of driving on the left, in an unfamiliar vehicle with the steering wheel on the right. It may be clear enough, in ad vance of the trip, that the combination of jet lag, fatigue, and unfa miliar task environment is potentially capable of producing a degra dation of driving performance. It may also be clear that “being careful”- which in this case means deliberately attempting to rely less on tacitly known skill-is likely to be at least partially effective as a compensating factor. But the problem of assessing the weight of these considerations, for the purpose of deciding whether the plan is acceptable or not, is intractable because of the tacit basis of driving skill. A full conscious override of habitual response is not possible, and if it were it would mean the abandonment, not the effective adaptation, of driving skill. The planner might reflect that the problem is surely not that serious; the muscular coordination aspect of controlling the vehicle will not require much attention. On the other hand, those muscular responses are tightly linked to visual cues, and the cues do not have their accustomed import. Habitual responses will be modified and the American driver will ilget the hang of i t” after a while, but it is hard to say how much experience will be needed or what risk levels might be involved in acquiring it. There is thus a significant degree of ambiguity about whether an American driver, driving for the first time lion the wrong side of the road,” knows how to drive or not. The ambiguity is partly a matter of uncertainty concerning the fate of the individual driver, and partly a reflection of the fact that the phrase “knows how to drive” papers over many significant distinctions.
Of course, if the American driver never goes to England, he may never directly confront this particular illustration of the ambiguous scope of skills. If he goes with sufficient frequency , he may develop a driving-on-the-Ieft subskill that is as much an integrated, tacitly known part of his overall driving skill as the ability to adjust to dense city traffic after coming off the ramp of a relatively uncrowded freeway . It is the differences between the environment in which a skill (and associated terminology) is developed and a relatively novel environment in which it is exercised that highlight its operational (and semantic) ambiguities . A fully static world would never pose the problem of using relatively concise language to consider the matching of complex skills with novel, complex task environments. The matches would all have been made, and could be counted on to work precisely as well in the future as they had in the past . But the real world is not static.
Source: Nelson Richard R., Winter Sidney G. (1985), An Evolutionary Theory of Economic Change, Belknap Press: An Imprint of Harvard University Press.