A Developmental History of Training
in the United States and Europe

Deborah Alpert Sleight
Michigan State University

December, 1993

Since prehistoric times the jobs of adults have been taught to children to prepare them for adulthood. Throughout the centuries societies have changed, have become more complex. The kinds of work to be done, the skills needed, and the tools used to do the work have also changed. In order to manage these changes in the complexity, volume, and content of work, job training also evolved. This paper describes how and why job training changed as work changed.

When humans created artifacts, the need for teaching others how to use those artifacts became necessary. Teaching others to use a tool to perform a task was one of the first goals of training. "As man invented tools, weapons, clothing, shelter, and language, the need for training became an essential ingredient in the march of civilization" (Steinmetz, 1976, p. 1-3). As the artifacts became more complex, different ways of training were developed to be more effective and efficient. "Instructional practices were developed that served the needs of the times, evolving into accepted instructional paradigms" (Brethower and Smalley, 1992, p. 26). These different practices were developed at different times, and some of them changed through the years, but all of them are used today, depending on the training need and situation.

Training is different from education. Training teaches the learner how to do a specific task, such as running a machine, or making a shirt. Education is instruction in the more general knowledge of the society, such as the history of the society, or knowledge of mathematics. As societies developed, there accumulated more knowledge than people could pick up on their own or learn informally from others. At some point it became necessary to formally educate young people in the amassed knowledge of their society in order to help them function in that society.

Antiquity: On-The-Job Training

On-the-job training, sometimes called direct instruction (or sit-by-me training in England; King, 1964, p. xvii), is the earliest kind of training. On-the-job training (OJT) is a face-to-face, one-on-one kind of training at the job site, where someone who knows how to do a task shows another how to do it. In antiquity, the kind of work that people did was mainly unskilled or semiskilled work not requiring specialized knowledge. Parents or other members of the group usually knew how to do all the jobs necessary for survival, and could pass their knowledge on to the children through direct instruction.

OJT was used in antiquity because it did not require learners to be able to read or write. Even after writing systems were developed, most peasants and craftsmen could not read or write. Therefore a type of training where one person showed another how to do a task was necessary. The type of work most people did was farming or making crafts, and the tools used were fairly simple. In addition, the volume of production was low, so that only a few artisans at a time needed to be trained in order to handle the work.

On-the-job training is still used today. In fact, it is probably the most popular method of training because at its minimal level it requires only a person who knows how to do the task, and the tools the person uses to do the task. The company doesn't have to arrange for special training other than to assign an experienced worker to train an inexperienced one. It may not be the most effective or the most efficient method, but it is the easiest to arrange.

The advantages of on-the-job training include instant feedback about what the learner is doing right or wrong, allowing correction of the erroneous action immediately. Because the training takes place on the job, it is realistic, therefore no transfer of learning is required. It is inexpensive because no special equipment is needed other than what is normally used on the job.

On the other hand, on-the-job training takes the trainer and materials out of production for the duration of the training time, and requires many experienced trainers--depending on the number of trainees--since the trainer is usually training one person at a time. However, if the volume of production is low, so that the personnel and equipment are available for on-the-job training, then it is the most economical method of training to use. It may not be the most efficient at teaching a task, however; this depends on the learner, the job to be learned, and the teaching abilities of the trainer.

The Middle Ages: Apprenticeship

Although apprenticeships had begun much earlier (according to Steinmetz (1976), rules for governing apprenticeships were included in the Code of Hammurabi in 2100 B. C.), they became more widespread during the Middle Ages. As tools became more complex, and the required knowledge and skills to use them became more specialized, parents or townspeople could no longer teach their children everything. Some children were then apprenticed to craftsmen who had the specialized skills and tools for a particular trade. In exchange for work, the craftsman would teach the child the craft at which he was an expert. "Apprentices usually lived with the master and received practically no pay, except maintenance and training" (Steinmetz, 1976, p. 1-4).

During the Middle Ages artisans produced their wares one at a time, which was sufficient to meet the demand. Because the skills are more specialized and there are more of them to master, apprenticeships last longer than on-the-job training, frequently for years. Apprenticeship continues today in the United States, and is, according to Patterson (1942):

...essentially a program of training beginners in the skilled trades, and that by an all-round skilled trade is meant one that takes at least two years to learn. It does not, properly speaking, apply to specialized one- or two-machine jobs...He [the successful apprentice] can adapt himself to different types of machines, and while he might not at first be quite so productive, he soon becomes proficient. Furthermore, he is not limited in the range of his adaptability. It is, then, to this latter all-round type of work that the term apprenticeship properly applies...an apprentice is not the same as a learner.' The latter term applies in connection with occupations which, in the opinion of the industry concerned, require a relatively short time to learn (pp. 136-137).

Modern apprenticeship principles protect both the master and the apprentice in terms of spelling out what work is to be done, how training is to be accomplished, how long the apprenticeship is to continue, and under what conditions it is to be carried out (Patterson, 1942, pp. 138-147).

The Industrial Revolution: The Classroom

It wasn't until the Industrial Revolution that training changed much from on-the-job and apprenticeships. In the 1800s factory schools were created, in which workers were trained in classrooms within the factory walls. In 1872 Hoe and Company, a manufacturer of printing presses in New York City, "had such a volume of business that it was necessary to establish a factory school to train machinists. The old-style apprentice system was inadequate" (Steinmetz, 1976, pp. 1-6). Classrooms by that time had become the customary places of education, and were simply recreated in the factories, with the classroom subjects being how to do the tasks required in the factory.

Why did factories turn to classroom training at that time? The machines of the Industrial Revolution greatly increased the ability of the factory to produce concrete goods quickly and cheaply, so more workers were needed to run the machines. The factory owners wanted the workers trained quickly because there was a large demand for the produced goods. Since the machines were much more complicated than the tools of the agrarian society of the past, and training needed to be accomplished quickly, the training methods of the past were inadequate.

Whenever it has become vital for a company to train large numbers of men on machine operations, especially on today's complicated machines, previously used methods have frequently proved inadequate. Most of the training programs conducted by industry were established to handle normal employment" (Smith, 1942, p. 59).

Unlike on-the-job training and apprenticeship, classroom training could train many workers at the same time and under a single trainer, so fewer trainers were needed. Learning away from the job kept distractions from the production floor at a minimum and did not take equipment out of production for use in training.

Classroom training may have been economically feasible, but it had some drawbacks that on-the-job training did not. Since workers were now learning away from the job, they had to remember not only what they had learned in the classroom until they could get onto the floor of the production line, but they also had to transfer what they had learned abstractly in a classroom onto a real machine in a real work environment. In the classroom the worker was now one student among many, and had to learn at the teacher's pace. Feedback was given less often, and sometimes not immediately after the learning. It was more difficult to ask questions without disturbing the rest of the class and without perhaps appearing stupid for asking the question.

Classroom teaching, whether by conventional lecture or by novel and exciting stimulus media, invariably addresses a hypothetical average student who might, in fact, not exist. The rapid learner is held back, and the slow learner--who might, nonetheless be a good student--is dragged forward too quickly. Few students have a chance to respond in any given session, and the teacher might favor those who are most likely to give gratifying answers (Cook and Mechner, 1962, p. 5).

Vestibule Training

A method of training was developed around 1900 that combined the benefits of the classroom with the benefits of on-the-job training and was a popular form of training in both world wars. This was called vestibule training, or "near-the-job" training. In vestibule training, a classroom or training room, located as close as conditions permit to the department for which the worker is being trained, is furnished with the same machines as are used in production. There are six to ten workers per trainer, who is a skilled worker or supervisor from the company (Smith, 1942, p. 62).

Vestibule training creates a miniature of the department for which the training program is carried on. It utilizes machinery similar to that in operation on the production floor. Qualified instructors, usually highly skilled operators or supervisors, are provided to conduct the program in this special section. Here the new employees are given a course of training in the particular machines they will be required to use and on the exact work they will do when they become a part of the regular production force (Smith, 1942, p. 59).

The advantages of vestibule training are many. The workers are trained as if on the job, but, according to Smith, "this training [vestibule training] will not interfere with the more vital task of producing, but will still serve to equip these needed employees for their jobs" (Smith, 1942, p. 60). Workers do not have to transfer their knowledge from a classroom to the work floor, and there are few enough workers per trainer that trainees may receive immediate feedback and be able to ask questions more easily than in a classroom. And finally, accidents are minimized because new workers' nervousness about working with unfamiliar machines is abated through simulated production.

There are some disadvantages to vestibule training. Vestibule training is expensive, because it duplicates the production line. Therefore it is best used when large numbers of workers need to be trained quickly, and for unskilled or semiskilled tasks that do not require long training periods, says Hardman (1963, p. 29). Because of the equipment and space requirements, numerous operations and machines cannot be duplicated for this training, particularly long-cycle operations and bulky machinery. Vestibule training is also "...very costly because full-time instructors are necessary, space must be allotted, and machinery must be either taken from production or purchased for this particular purpose (large initial expenditure and continued sizable operating costs). However, considering the savings achieved and the value of this type of training, these costs are extremely insignificant" (Smith, 1942, p. 66).

The vestibule training method, dusted off and brought up to date, has demonstrated great possibilities in helping industry meet the problem of training new employees for war production. It does not attempt to replace other methods of training but merely to supplement them where they are weakest. Its value is greatest where there are large numbers of men to be trained on unskilled or semiskilled machines or mass-production operations in a relatively short period of time. The present methods must still be used for the training of skilled help or of smaller numbers. But vestibule training, properly utilized, will fill today's urgent needs (Smith, 1942, p. 66-67).

The World Wars: Systematic Training

The world wars brought an urgent need to train large numbers of defense workers quickly to fill the huge demand for products. "With the growing complexity and size of factories, expanding markets that exerted a strong demand for an increase in the volume of production, and a rising engineering profession, there emerged a new and pressing concern to systematize the administration, control, coordination, and planning of factory work" (Zuboff, 1984, p. 41).

Many of the experienced workers were enlisting, so not only were workers needed to fill those positions, but to fill numerous new positions to meet the demand for goods. To this end, methods were sought that would train workers more quickly and thoroughly than previous methods had.

Around 1915 Frederick Taylor conceived of a method for shortening the amount of time a task took by studying workers doing the task and removing "non-productive time." He called this method Scientific Management. Frank and Lillian Gilbreth added to Taylor's work by studying workers' movements on the job and suggesting ways to simplify the job and minimize the number of movements (Westgaard, 1993, p. 87).

During World War I a system was proposed for on-the-job training that would provide more consistent and efficient training in order to more quickly produce trained workers. In 1917, in response to the need of 450,000 new workers by the Emergency Fleet Corporation of the U.S. Shipping Board, Charles R. Allen developed a way of training shipbuilders, which involved four steps: 1) SHOW or prepare, 2) TELL or present, 3) DO or apply, and 4) CHECK or inspect. From Allen's work and from research done by the Army during World War I several principles of training instruction were developed (McCord, 1976, p. 32-6):

These principles of instruction were used in industrial training, but no method of systematic training appeared until World War II. According to McCord (1976), "... some discussion did take place as the result of Allen's work and although a few books on how to instruct did appear during the 20s and 30s, not until the need for defense production became serious in 1940 were steps taken to develop a method of instruction" (p. 32-37).

In 1925 Gardiner (1925) in Practical Foremanship described how a foreman should teach his "green" men to do the job, employing the principles of industrial instruction and the studies by Taylor and the Gilbreths on job simplification. "There is a one best way to do a job, and it is assumed that a foreman knows the one best way to do each job in his department" (p. 122).

In order to know the one best way to do a job, the foreman first had to analyze and simplify the job, then teach each separate operation. He was to pass on the little hints that experience teaches and the common-sense facts of the job, as well as the whys and wherefores of the job (Gardiner, 1925, pp. 122-123).

Instead of assuming that workers had to be pushed to produce, he was to set the learner at ease about the work and his fears of awkwardness and stupidity, and was to make difficulties appear small. He was to teach one thing at a time and make sure that the learner understood the first point before teaching the second. He was to ask questions of the learner that showed if the learner actually understood what he had learned, and was to follow up on the employee's progress (Gardiner, 1925, pp. 123-125).

During World War II, the need for a method of fast and efficient training became urgent. Training Within Industry, an advisory service formed by the National Defense Advisory Commission, developed the systematic on-the-job training method called JIT (Job Instruction Training). There were four steps to this method (Hardman, 1963, p. 97):

According to Hardman (1963), JIT is useful primarily in teaching manipulative skills, but could be disappointing with more complicated tasks. An advantage of JIT is its flexibility; one variation has the teacher and learner switch roles. In order to be successful JIT requires adequate supervision and uniform application throughout the plant.

After World War II: Individualized Instruction

During World War II training departments had become established in many companies. Business was booming as the world rebuilt itself. Companies wanted their workers trained, but as usual they wanted a more efficient, less expensive method of training. Upon the concept of job analysis and the behaviorism of B.F. Skinner was created the concept of individualized automated instruction. Just as in industrial work, training itself was about to become automated.

Individualized instruction in essence replaces the teacher with systematic or programmed materials. Programmed materials are instruction that has been divided into small steps which are easily understood by the learner. After each step is required an active response by the learner in the form of answering a question, drawing a graph, solving a problem, and so on. Immediate feedback is given after each response. Individualized instruction can be print-based, computer-based, or can use other media as long as the instruction is based on the concepts listed above. Although logical subject matter is easier to individualize into programmed materials, Lysaught (1962) had not found any subject matter that could not be programmed.

Sidney Pressey's testing device in 1926 and B.F. Skinner's behavioralist theory of learning in 1954 were the basis for the new programmed instruction, and the teaching machines in which it could be automated. Skinner's model of programmed learning was linear, in that the author of the materials decided what step to present next, and that step was presented, no matter what the learner wanted. According to Westgaard (1993), "many programmed learning materials are reminiscent of the Gilbreths' thinking" in terms of designing work so there are as few as possible decisions to be made by the worker, or branches by the program (p. 88).

But Robert Glaser "had spoken of the desirability of having a provision for skipping' items in an otherwise sequential program, thus permitting the rapid learner to by-pass items which might seem redundant to him, but which were desirable for those students encountering difficulty with the program" (Lysaught, 1962, pp. 30-31). Norman Crowder developed what he called "intrinsic" programming, in which the learner's possible responses are multiple choice, and the program branches according to the response chosen (Lysaught, 1962, pp. 30-31). In this way students could skip steps they already knew, or study remedial material on information already presented.

Individualized instruction was originally presented in book form, and sometimes still is. In order to prevent students from looking at the answers in the book ahead of time, some individualized instruction was automated by inserting it into a teaching machine. Teaching machines are "devices that house, display, and present printed programmed instruction...A display aperture exposes each instructional frame, and a second aperture provides a space for the response. Feedback is given when the program is advanced through actuation of a lever, knob, or button, and the correct answer comes to view. It is probably best that the frame and the learner's answer both remain in view when the feedback is presented" (Cook and Mechner, 1962, p. 11). Although teaching machines could score selected answers, they could not score constructed answers. This problem has continued to this day in computer-based training, although some progress has been made in getting the computer to be able to recognize and evaluate constructed answers.

The advantages of individualized instruction are many: it allows a learner to work through the material at his or her own pace, and provides privacy in which to make mistakes and correct them. Individualized instruction reduces learning time, produces a low error rate in the learner, improves learning through immediate feedback, and provides consistent teaching.

It reduces the amount of training needed to do a job by "incorporating automatic, on-line feedback signals...Thus, the magnitude of the task assigned to training is reduced" (Cook and Mechner, 1962, p. 13). Individualized instruction can also reduce the aptitude needed to do a job. "The increased instructional effectiveness that programing will bring might well alter the strategic role of selection tests and broaden the base of recruitment to a wider population range. Allocation strategies for already selected men similarly might become more flexible" (Cook and Mechner, 1962, p. 12).

As with anything, there are disadvantages, too. Individualized instruction is expensive to produce because it takes a long time and requires skilled people to analyze the subject matter and prepare the materials. It teaches only what the author includes. Finally, it is usually learned off-task, requiring the learner to transfer that learning to the job.

Modern Types of Individualized Instruction

Individualized instruction is the basis for computer-based training, which is used extensively in many companies today. Computer-based training (CBT) provides individualized instruction via the medium of computers. It was a natural progression from printed individualized instruction and teaching machines to the computer with its speed, branching capability and visual display. CBT was developed in the 1970s on mainframe computers, when the increasing capability and capacity of computers and their decreasing cost made CBT technologically feasible (Gery, 1987).

The definition of CBT is very close to the definition of individualized training used earlier: "an interactive learning experience between a learner and a computer in which the computer provides the majority of the stimulus, the learner must respond, and the computer analyzes the response and provides feedback to the learner" (Gery, 1987, p. 6). The computer can go beyond the original print- or teaching machine-based instruction because it has more capability for automatic branching and scoring. CBT can be used to teach almost any kind of subject matter, as long as the design of the instruction is sound. Its multimedia function has added the capability of displaying information in audio, graphic, and motion video form, which makes the teaching of skills and processes more effective than if only text were used. Individualized instruction delivered over the World Wide Web is a further development of computer-based training. Web-based instruction can be used with any type of computer that can access the internet and that has web browser software.

A specialized form of computer-based training is called embedded training. Developed by and used primarily in the military, embedded training "involves incorporating training functions, either in whole or in part, into an operational system" such as air-to-air engagement and air combat maneuvering (Jorgensen, 1991, p. 947). Since the operational system is computer-based, the training function must be computer-based as well. The military uses embedded training to teach tasks that are critical and perishable. "Perishability was defined as the extent to which a particular skill requirement...tends to decay without periodic reinforced practice. Criticality has its usual definition of importance to mission success'" (p. 948). The Navy uses embedded training to train its personnel to proficiency on various systems. "Certainly, an operational system should not be used for training already accomplished in schools or elsewhere before personnel are assigned to the job. However, if skills are perishable, that is they are lost over time, repeated training may be required to maintain required performance levels, and ET can provide that" (p. 951).

The study reported by Jorgensen listed several potential benefits of embedded training: more opportunities for realistic training and feedback; increased availability and accessibility of training in operational units; and more opportunities for conducting training in tasks where training is currently limited by costs, safety, or air traffic restrictions" (p. 947). The disadvantages include trying to maintain it and keep its documentation up to date, problems common to any computer-based training.

The End of the 20th Century: Job Support

The adaptive process which we call "learning" is obviously fundamental to any understanding of the nature of training. -- King (1964, p. xix)

All training methods so far have had as their basic goal that the workers learn what is presented. That is why the quotation from King at the beginning of this section was true when he wrote it in 1964, and why it is true as long as the work is fairly stable and concrete, as it was until the advent of the knowledge society. The knowledge society is a society where "...the use of knowledge and information dominates work and employs the largest proportion of the labor force. The distinguishing feature of a knowledge society is that it emphasizes intellectual work more than manual work--the mind more than the hands" (Newstrom and Davis, 1993, p. 333).

At the end of the 20th century we are in the middle of the information revolution, where the amount of knowledge required to do a job is growing steadily, work procedures may change rapidly, and frequently the goods produced are not concrete, but abstract, that is, information. In such a work environment, learning the changing procedures and all the information needed to do the job is practically impossible. A new form of training is needed to manage such a change in work.

Job support is a form of training that does not require the worker to learn, but simply to know where to find the changing information they need. Job support tools contain step-by-step instructions on how to do a task, they are designed to be used on or just before the task, and the worker does not need to learn the steps as long as he or she knows where to look to find the specific job support tool needed at the time. Instead of learning the information needed to do a task long before the task is to be done, as in a classroom, the worker can simply follow the instructions on the job support tool at the time the instructions are needed. In fact, job support tools have transformed from training tools into job tools.

A well-known type of job support tool that has been in use for decades is the job aid, which traces its modern roots to the JIT method used in World War II. Chalupsky and Kopf (1967) define a job aid as "...documentation or auxiliary information storage and presentation devices used on the job to provide instructions and reference information. They are designed to facilitate employee performance through direct usage either during the course of task performance or perhaps immediately prier to beginning a task" (p. 7).

The job aid began as a printed card that contained step-by-step instructions for performing a specific task. The worker did not need to memorize the steps because they are remembered for him on the card. "Job aids are considered instructional interventions because they also mediate knowledge/skills problems. However, job aids are not really intended to produce learnin--they substitute for learning. Learning that does occur as a result of using the job aid (surely considerable at times) is incidental" (Ruyle, 1990, p. 4).

Not every task is suitable for "learning substitution," however. Jobs that are done frequently, so that the worker learns the steps anyway, need no support. Jobs that require quick thinking and action, such as emergency procedures in an airplane must be learned to mastery so they may be performed quickly and effectively when necessary because there may not be time to read and follow a step-by-step job support tool. But jobs that are done infrequently or must be done absolutely correctly or in a particular order every time are a prime candidate for job support. An example of such a job support tool is the flight departure checklist. The pilot and copilot of a plane likely remember what steps to take to prepare for takeoff. But what if they forgot one step? What if they reversed the order of two crucial steps? Many people could die. Because the steps must all be done, and done in the correct order, pilots use a departure checklist before taking off.

Another example of a job whose performance can be improved by support is patient care. "The critical importance of the patient care tasks and the need for extreme accuracy and attention to detail, coupled with problems of high personnel turnover and the need to improve the efficiency of personnel utilization, further argue for consideration of the patient care areas as a test bed' for job aid demonstration" (Chalupsky and Kopf, 1967, p. 66).

Another type of job support tool is an expert system, a computer program that has been given rules to follow which contain the knowledge of experts in a given area. A worker doing a task in that area can ask the expert system a question and receive advice as if from an expert.

According to Chalupsky and Kopf (1967), the advantages of job support tools are reduction in training time; lowering of selection standards; improved quality and reduction in job errors; increased reliability of job performance; greater productivity; increased scope and complexity of tasks performed" (p. 7).

There are disadvantages with job support tools. They can be difficult to follow if there are distracting interruptions. They are not useful when quick responses are required because the worker has to read, understand, and follow them at the time of the task. They do not support complex tasks very well because they become too complex to be of much use. "They also lose their effectiveness when they pile up to the point that the individual has trouble keeping track of them" (Carr, 1992, p. 179). In addition, some of the advantages may be seen as disadvantages from a different point of view. If the use of job support tools can lower personnel selection standards, the tools may be deskilling jobs, taking the cognition effort out of the job and reducing it to a simple mechanical exercise. Effective job performance aids, by reducing the information retention load in a job, can ideally serve as an alternative to the continual expansion or graining that must otherwise result as jobs become more complex or as the input quality of the work force decreases (Chalupsky and Kopf, 1967).

The Information Revolution and Beyond: Integrated Performance Support

As work changed, training changed to support that work. The changes in training methods appear to have been caused by changes in the kind of work being done. These changes in work involved the kind of goods produced (concrete or abstract), the kind of skills the job required (unskilled, semiskilled, or skilled), and the size of the demand for the goods. The changes in work were caused by war, by advances in technology, or by other changes in societies.

Those in charge of workers are always looking for ways to speed up training, ways to make it more economical, efficient and effective, ways to minimize the need for it. Just as many scientific discoveries can be used constructively or destructively, so new methods of training can be helpful or destructive. Frederick Taylor's Scientific Management concept and the Gilbreths' time and motion studies helped simplify and speed up work, but they also dehumanized the workplace and the worker by doing so.

Workers, for Taylor, were simply parts of a mechanical process. Like tools and machines, they could be replaced when they wore out. His goal was to reduce the skill needs of workers to as close to zero as possible and, thereby, eliminate drains on productivity like the need to train workers or retain them when they began to ask for higher wages (Westgaard, 1993).

Frank and Lillian Gilbreth "strove to increase linearity and reduce branching in a process. The goal was, in effect, to develop step-by-step processes driven by their own inertia. In the ideal process,' workers had no choices. Workers had to do the right thing at the right time...The key to success was to identify and eliminate elements (usually human tendencies) that blocked the process. The goal was to remove any decision points in the process" (Westgaard, 1993, p. 87). Donald Norman says that what is wrong with the design of technology is that it "requires people to behave in machine-centered ways, ways for which people are not well suited" (Norman, 1993, p. 11). This is what Taylor and the Gilbreths were attempting to do, to make workers part of the machines they were using.

According to Chalupsky and Kopf (1967), job aids "...can compensate for lack of training and experience..." (p. iii) and perhaps for basic aptitude deficiencies. If this is the case, does job support deskill the work, making it less satisfying to do? Is this always the case with new training methods and technology, that they are more economical and effective because they reduce the reliance on human brains and skill? If so, what will be the likely result of such a progress? Zuboff (1984) notes that "in diminishing the role of the worker's body in the labor process, industrial technology has also tended to diminish the importance of the worker. In creating jobs that require less human effort, industrial technology has also been used to create jobs that require less human talent" (p. 22).

Performance support technology seems to present two sides of the same problem: 1) the potential deskilling of work, and 2) the potential capability of compensating for basic aptitude deficiencies. Workers that can do the present form of the job may not like the job if it is less of a challenge to do. On the other hand, workers that cannot do the present job, but need work, may be relieved to find work they can do. With regard to the deskilling of work, Zuboff says that information technology, by changing work activities to the abstract, can "imply reconstruction of knowledge of a different sort."

Information technology, however, does have the potential to redirect the historical trajectory of automation [deskilling of jobs and easing of physical labor]...The new technology signals the transposition of work activities to the abstract domain of information. Work' becomes the manipulation of symbols, and when this occurs, the nature of skill is redefined. The application of technology that preserves the body may no longer imply the destruction of knowledge; instead it may imply the reconstruction of knowledge of a different sort" (Zuboff, 1984, p. 23).

This different sort of knowledge, Zuboff thinks, will be intellective skill, "...a shift away from physical cues, toward sense-making based more exclusively upon abstract cues; explicit inferential reasoning used both inductively and deductively; and procedural, systemic thinking" (p. 95). Workers' essential skills will likely change as the sort of knowledge required changes, but there will always be essential skills that a machine cannot do.

Instead of replacing workers' skill, Carr calls instead to preserve the workers' core competence. "Neither job aids nor any other form of training should ever reduce workers' competence in any core functions of their jobs. Conversely, whenever a firm can reduce the time and attention that workers have to pay to their inessential duties it should make every effort to do so (Carr, 1992, p. 179).

Because performance support systems can use expert systems to make decisions formerly made by humans, there's a real temptation to use them to replace the skills of workers...If you attempt to substitute expert system-based performance support systems for essential employee skills, you'll reduce workers' core competence_and thus the overall core competence of your firm. On the other hand, if you use these same systems to take over the often irritating inessential tasks, you'll preserve core competence. You'll also make your workers happy" (Carr, 1992, p. 181).

If management were to believe and follow Carr's exhortation and preserve core competence while using support tools to "make it easy for individuals to do the inessential parts of their jobs--so they can concentrate on the essential or core tasks" (p. 178), then the problem becomes to identify what is essential and what is inessential about a job. I would guess that an individual worker could do this fairly easily for his or her own job. But what happens when the job changes because of technology? How will the worker identify the new skills essential for the job then?

One of the research recommendations of Chalupsky and Kopf (1967), authors of a study of job aids in the 1960s, was to learn the potential contribution of job aids for reducing basic aptitude requirements. "...depending upon their design and usage, job performance aids can compensate for lack of training and experience, improve job quality, and increase productivity. Evidence is generally lacking concerning the extent to which job performance aids can compensate for basic aptitude deficiencies. The impact of job aids is particularly apparent in complex or lengthy tasks, diminishing in effectiveness as tasks are subdivided into short cycle, repetitive operations (p. iii). Performance support systems may be able to reduce basic aptitude requirements. But is this all bad? People with less aptitude and education have to make a living, too.

What jobs are being created now and will be created, and what kinds of skills will they require? What kinds of jobs are disappearing, and what kinds of skills did they require? Do workers feel a need for technology that would make jobs available to them that currently require too much education and experience? There are many questions still to be answered and many questions not yet asked.


Brethower, D. & Smalley, K. (1992). An instructional paradigm for training in the 21st century. Performance and Instruction, May/June 1992, 26-31. Washington, D.C.: National Society for Performance and Instruction.

Carr, C. (1992). Smart Training. New York: McGraw-Hill.

Chalupsky, A. B. & Kopf, T. J. (1967). Job performance aids and their impact on manpower utilization. U.S. Dept. of Labor: Office of Manpower Policy, Evaluation and Research. Palo Alto, CA: Philco-Ford Corporation.

Clark, R. (1992). EPSS--Look before you leap: Cautions about applications of electronic performance support systems. Performance and Instruction, May/June 1992. Washington, D.C.: National Society for Performance and Instruction.

Cook, D. & Mechner, F. (1962). Fundamentals of programed instruction. In S. Margulies & L. Eigen (Eds.), Applied Programed Instruction. London: John Wiley and Sons, Inc.

DePhillips, F. A., Berlin, W. M. & Cribbin, J. J. (1960). Management of Training Programs. Homewood, IL: Richard D. Irwin, Inc.

Dodd, A.E . & J. O. Rice (Eds.) (1942). How to Train Workers for War Industries. New York: McGraw-Hill Book Company.

Gardiner, G. L. (1925). Practical Foremanship. New York: McGraw-Hill Book Company, Inc.

Gery, G. (1987). Making CBT Happen. Boston: Weingarten Publications.

Gery, G. (1991). Electronic Performance Support Systems. Boston: Weingarten Publications.

Goldberg, I. (1962). An introduction to programed instruction. In S. Margulies & L. Eigen (Eds.), Applied Programed Instruction. London: John Wiley and Sons, Inc.

Hardman, W. E. (1963). In-plant Training. Waterford, CT: Prentice-Hall, Inc.

Heaford, John M. (1983). The Myth of the Learning Machine: The Theory and Practice of Computer Based Training. Cheshire, UK: Sigma Technical Press.

Jorgensen, W. F. (1991). Embedded training in a system. In IEEE National Aerospace and Electronics Conference, 2, 947-953. Piscataway, NJ: IEEE.

King, D. (1964). Training within the Organization. London: Tavistock Publications.

Leymann, H. & Kornbluh, H. (1989). Socialization and Learning at Work. Brookfield, VT: Gower Publishing Company.

Lysaught, J. (1962). Programed learning and teaching machines in industrial training. In S. Margulies & L. Eigen (Eds.), Applied Programed Instruction. London: John Wiley and Sons, Inc.

McCord, B. (1976). Job instruction. In R. L. Craig & L. R. Bittel (Eds.) Training and Development Handbook. New York: McGraw-Hill Book Company.

McGehee, W. & Thayer, P. W. (1961). Training in Business and Industry. New York: John Wiley and Sons, Inc.

Newstrom, J. W. & Davis, K. (1993). Organizational Behavior: Human Behavior at Work, 9th edition. New York: McGraw-Hill, Inc.

Norman, D. (1993). Things that Make Us Smart: Defending Human Attributes in the Age of the Machine. Reading, MA: Addison-Wesley Publishing Company.

Patterson, W. O. (1942). Apprenticeship principles for personnel managers. In A. E. Dodd & J. O. Rice (Eds.), How to Train Workers for War Industries. New York: Harper and Brothers Publishers.

Ruyle, K. E. (1990). Developing Conventional and Intelligent Job Aids: A Case Study. Unpublished dissertation. Corvallis, OR: Oregon State University.

Scanland, D. von K. & Scanland, W. (1981). Automated job performance aids: Concept and state of the art. NSPI Journal. Washington, D.C.: National Society for Performance and Instruction.

Smith, L. J. (1942). Vestibule training for today's needs. In (A. E. Dodd & J. O. Rice, Eds.), How to Train Workers for War Industries. New York: Harper and Brothers Publishers.

Steinmetz, C. S. (1976). The history of training. In (R. L. Craig, Ed.) Training and Development Handbook: A Guide to Human Resource Development, 2nd Ed. Sponsored by the American Society for Training and Development. New York: McGraw-Hill.

Westgaard, O. (1993). Allow me to introduce...Frederick M. Taylor and the Gilbreths. Performance Improvement Quarterly, 6(1), 86-90. Learning Systems Institute, Florida State University, in cooperation with the National Society for Performance and Instruction, Washington, D.C.

Zuboff, S. (1984). In the Age of the Smart Machine: The Future of Work and Power. New York: Basic Books, Inc., Publishers.

(c) Deborah Alpert Sleight, 1993
Permission is given to reprint for non-profit use providing credit is given.

Deborah Alpert Sleight
Michigan State University
East Lansing, MI 48824