Between craft and system

explaining project-based enterprise in the United States, circa 1910

Benjamin W. Pinney, Massachusetts Institute of Technology

The histories of project and program management in the United States have most often been traced to the 1950s and to Cold War contracting and systems management (Hughes, 1998; Bugos, 1993; Johnson, 1997; among others). The late 1950s and early 1960s witnessed the popularization of the term “project management” and the broadcasting of the idea by academics, managers, and engineers. Coalescing around newly invented network scheduling techniques such as Critical Path Method (CPM) and Program Evaluation and Review Technique (PERT) and disseminated through military contracting requirements, project management enjoyed a heyday in the 1960s. Over the course of that decade private enterprises wove projects and programs into their hierarchies to create matrix organizations and consulting firms set up practices geared to helping businesses “projectize.”

While some of the faddishness that attended project management in the 1960s later dropped away, thinking about and organizing for work in terms of projects now pervades language and practice. The last ten years have seen a renewal of interest in project management and the sociology of projects. An article in PM Network (McCowan, 1999) calls manufacturing “the final frontier” for project management as product cycles accelerate and the last bulwarks of mass production give way. Paul Dinsmore (1999) describes firms as no longer rigid hierarchies but fluid aggregations of projects. In common, these and many writings on the “new economy” herald a world in which people experience work less in terms of day-to-day routines and corporate ladders than in terms of irregular work on temporary tasks and careers built on networks of inter- and extra-firm personal and professional allegiances.

Yet, for some kinds of business and for some kinds of work the present is as it ever has been. Discussing shifting social relations between employers and employed, The Economist (2000) has remarked: “What on earth is happening? The broad answer is that it depends on where you live and what you do. The ‘insecurity’ that is suddenly so frightening to a sarariman in Tokyo has long been normal for a construction worker in Dallas …”

Studies of project management often acknowledge such connections. Construction work provides the “archetypal context” for the application of project management (Lundin & Midler, 1998, p. 4) and sociologists have offered rich considerations of the structural kinships between the organization of craft and professional labor in different industries (Steiger & Form, 1991; Stinchcombe, 1959, among others). However, the historical connections among these occupations have been little studied and, with notable exceptions (Boutinet 1996; Morris, 1994), the history of project management before 1950 is almost unexplored.

In this paper I argue that the formal development of what is now called project management began in the late nineteenth century, in tandem with the general development of systematic management. The first clear articulations of projects as a distinct form of production—and one requiring distinct means of management—appeared in the writings of civil engineers during the late nineteenth century. The word “formal” is crucial here. What was new in this period was neither projects nor their management, but the need for and availability of written, codified methods of control. In this period, the scale and complexity of engineering work outgrew the capacities of informal communication networks and traditional craft-based working relations. Further, the speed of technological change and time pressures on individual projects exceeded the rate at which such older means of coordination could adapt. The contemporary development of accounting and management methods gave engineers new tools for organizing projects that answered many of these functional needs. At the same time, it was the development of factory systems that made projects into something different that needed explanation.

I next briefly discuss recent work in organizational sociology that helps guide my search for and interpretation of historical evidence. I then present several illustrative case studies drawn from engineers’ descriptions of their own work. Finally, I locate these cases against an overview of broader developments.


Recognizing the structural and historical continuities that link past practices to present-day project management is partly a matter of careful detective work. It is also a matter of translation. The neologisms and acronyms of the Cold War and, more recently, computer networks and software engineering obscure basic continuities with earlier ways of managing and working—the term “project management” itself only appears in periodical indices in the 1960s. Organizational sociology provides valuable translation tools. Of particular use is work that interprets organizations as arrangements for information processing. Building on studies done by Jay Galbraith (1969) and others, this approach examines organizations in terms of the formal and informal networks of communication used to get work done. Applied to craft and professional settings, this perspective crystallizes a fundamental difference between organization for repetitive work and organization for irregular tasks. Managers of repetitive work, at least in theory, can fully specify the processes and environment that shape outputs. In managing projects, the tasks required to meet the organization’s objective cannot be fully specified in advance because working conditions cannot be entirely known in advance. Arthur Stinchcombe (1990, p. 65) notes that in the building industry, “It is construction foremen who adjust the amount of labor the firm hires to the amount of work there is to do. It is construction workers as well as foremen or managers who coordinate the work of different crafts. There is no particular ‘buffering’ of the work of the skilled workers in their routines to take out the uncertainty; after all, the firm hired skilled construction workers in the first place because it was in an uncertain work environment.”

Where even modest time pressures exist, environmental uncertainty requires decentralization of decision-making. Though the separation of decision making from execution could hypothetically be applied even in the absence of a certain environment, it would be grossly inefficient. Moreover, builders—whether wielding hammers on a construction site or writing code in the cubicles of a software start-up— are not simply in uncertain environments, they are actively shaping and interacting with their environment. In other words, in looking for evidence of precedents for project management, the researcher needs to look for circumstances involving construction work—broadly interpreted.


The following four cases are drawn from a larger body of evidence that I have collected in looking for builders’ descriptions of how they organized for and accomplished their work. Together, they bridge a period of definition and for-malization for management practices generally. Individually, they track a chronological progression that marks both the increasing kinship between construction work and engineering in other fields and also the growing external pressures faced by individual managers.

W. Milnor Roberts and Railroad Engineering,1878

In 1878 the journal of the American Society of Civil Engineers published a talk by W. Milnor Roberts (Roberts, 1878). Born in 1810, Roberts had had an active career ranging from re-connoitering the transcontinental route of the Northern Pacific Railway to chairing a study of Philadelphia’s water supply. Addressing a meeting of his fellows, Roberts recounted a career that spanned both the transition from canals to railroads and that from informal to formal methods of management.

In 1831, Roberts joined the Pennsylvania-financed Allegheny Portage Railroad as an assistant and supervised the construction of several of that road’s inclined planes. As suggested by its name, the Allegheny was a short line connecting two waterways. These were separated by a ridge. American engineers of the time, following English precedent, considered locomotives impractical on any but the most level tracks and inclined planes served as the Portage Railroad’s equivalents to a canal’s locks. Along level sections horse drew carriages but where rises were required the animals were released and stationary steam engines hauled the train by ropes attached to the lead car. The business operation of this road was closely akin to a canal or toll highway in that the customer supplied the vehicle and the company provided only the way.

When the Allegheny opened in 1834, Roberts stayed on and had “the general charge and management of the running of the road … as well as the general superintendence of the working of the ten inclined planes” (204). The volume of traffic made this a complicated business: “any persons who chose could put cars on the track, and haul them, with their own horses, from either end to the foot of the first plane, and on the levels between the planes; the State merely transporting the cars up and down the inclines…” (204). As traffic increased, this “demanded almost constant attendance on the line, as questions daily arose, in the use of the road by various independent parties, which I was daily called upon to settle” (204). Recalling this business from the perspective of the late 1870s, Roberts could mark such a state of affairs as having since been overcome through technical and managerial innovations. Then, he noted, “the ‘telegraph’ was not yet born; and railroad experience was still a thing in embryo. The gigantic railroading of the present day had not then an existence even in dreams. Neither the ‘General Manager’ nor the ‘Receiver’ had yet been invented” (205). For Roberts, “engineering operations” embraced both building and operating the railroad.

Russell Robb and Electrical Engineering, 1909

In the spring of 1909, Russell Robb, an executive with the Stone & Webster Company of Boston, delivered three lectures at Harvard Business School, later collected in a privately published book, Lectures on Organization (1910). The first of his lectures, “Organization as Affected by Purpose and Conditions,” provided a series of examples from his own firm’s experience. He noted especially pressures arising from the cost of capital given that today’s construction was tomorrow’s production capacity. Every day cut from a construction schedule was a day that could be added to productive use. Where time pressures were great, “organization for such an undertaking will not be the same as for deliberate construction systematized in all details for lowest construction-cost” (7). He referred specifically to Stone & Webster’s experience building additions to power stations for the Boston Elevated Railway. The additional capacity had been needed to meet anticipated Christmas-season loads the next year: “The organization of a street railway could not economically provide for emergency construction of such magnitude, and even the engineering company that did the work found it necessary to make important modifications in its organization in order to secure dispatch, even at the risk of losing the greatest economy in details. It was necessary to place as much responsibility down along the line as possible, so that time would not be lost in referring questions for confirmation. It was necessary that the work be segregated, so that in functional departments there would be no question of ‘right of way.’ The specialization was in the direction of ‘Boston Elevated work,’ not in the direction of ‘electrical engineering,’ or ‘purchasing,’ or ‘drafting.’ A special smaller organization was set off for this particular piece of work” (8).

Robb went on to describe an organization in which the construction manager had had complete authority over a “an organization for dispatch” (8) that cut the time to put the stations into service in half relative to work done in the regular functional divisions of the firm.

Robb’s overall message was that system and order were admirable goals and necessary in an era in which competition, time pressures, and shrinking margins allowed little room for trial-and-error or rule-of-thumb approaches to business. He emphasized, however, the importance of not becoming enamored with any one approach or becoming “twisted with specialization” (15) and over-application of functional divisions. “Organizations are not difficult to sketch out on paper. They always have there a definite and workable look, as if nothing could escape a far-reaching arm that would pull all into the hopper somewhere; but in the working organization, unless it be the most simple, there is a constant call for the unscheduled cooperation, for the action that can be secured only through a genuine, lively, and loyal interest in the success of the whole undertaking. And it is interesting here to note that this sort of thing is often expressed by a simile that is the outcome of the activities of intercollegiate life, for we urge this spirit by urging ‘team play’” (22).

He concluded, “We begin to realize that there is an art of organizing that requires knowledge of aims, processes, men, and conditions, as well as the principles of organization” (22). Robb’s second lecture, “The Limits of Organization,” noted “the returns from increasing organization do not continue proportional to the effort, and limits are reached, beyond which one may well proceed with care” (24).

Ernest McCullough and Construction Management, 1915

In 1915, Sanford E. Thompson and William O. Lichtner presented a paper to the Western Society of Engineers on “Construction Management.” In this paper, they explained the practices of their firm in building formwork for concrete structures. Applying systematic management to construction required “the services of a new man,” they offered, a “production engineer” charged with “the treatment of the construction organization and methods in full detail” (115). The system to be applied by this new man involved minutely detailed tracking of every aspect of work, down to filing a separate “operation ticket” to cover cutting a piece of lumber into three pieces, with the dimensions of both the original and cut pieces of wood clearly spelled out. Thompson and Lichtner argued that the high administrative costs so incurred were more than compensated by savings in the time and materials of construction.

In the published discussion of Thompson and Sanford’s article, Ernest McCullough (1915), also a contractor, strongly objected to the limited relevance of their system. Under scientific management, he averred, “The productivity of the men employed by contractors has been increased by taking from men on the job the necessity of thinking beyond their task. The brains of the company are located in the head office, the convolutions of the brain being represented by masses of red tape, the gray matter being represented in white, yellow, green, red and other color sheets and cards. The job superintendent is an operator at a switchboard receiving and transmitting calls. Such highly developed systems are good, as we all know, for large organizations where the initial expense can be borne and a large overhead expense carried, which can be reduced to a small percent of the cost because of the large volume of business. But we need something better adapted to the sporadic contract undertaken by the individual, or the very small company, lacking capital and working in sections of the country where each job means a new organization. There must be considerable waste on such work, but as many contractors make some money, might it be possible to enable them to earn more and be more humane to their men by the introduction of common sense methods adapted to the understanding and financial ability of the average contractor?” (131–132).

For McCullough, such a system would be more about planning and less about paperwork. “Because a job does not bristle with hooks on which to hang blanks and is not plastered with special forms does not mean it is not systematically run” (134).

Speaking from his personal experience as a job-site superintendent, McCullough reviewed some of the factors that complicated the application of system to contract work. In his remarks, he was carefully sanguine about the desirability of efficiency in contracting; however, he made clear that irregularity of contracting made typical systems inefficient. Foremost in checking the value of system was the nature of the labor market. “The labor question is ever with us,” McCullough remarked, “local labor is seldom efficient and imported labor belongs to the ‘drifting class’” (133). Complicating the difficulty of finding and keeping good workers for McCullough was that hand-to-mouth contracting operations could not afford to pay workers—or sometimes even to keep equipment—between jobs. He concluded his remarks, “I believe in efficiency, and when system means efficiency I am for good system. When system means the pre-organization of a large staff at considerable initial expense, then I believe the efficiency experts have to revise their message to engineers—perhaps add a postscript” (135).

John Hardecker and Aircraft Manufacture, 1926

During World War One and the interwar period, aircraft, a novelty that began in experimental, one-off production, became a batch-produced good. But the rate of innovation in aircraft design and the underlying concern that the product be competitive in war limited practical production runs. Most aircraft were obsolete long before economies of mass production could be effected. Producing aircraft was a peculiarly hybrid process. The production run for a given model was both limited and subject to ongoing modification.

In 1926, John Hardecker (1926a, 1926b), an engineer with the Naval Aircraft Factory of Philadelphia, described that organization’s administrative structure in a pair of articles in the American Machinist: “The most flexible type of technical organization yet devised that can be successfully carried out in actual practice is the one in which a single engineer is given the task of ‘living’ with a job from start to finish so that at all times he is the one man in the organization who is most familiar with that particular job” (1926a, p. 641).

That “one man,” was a “project engineer.” Hardecker went on to note that this organization “lends itself readily to the manufacture of machinery, locomotives, automobiles, electrical equipment and, in fact, to any large plant making a number of mechanical or structural products” (1926a: 641). From this introduction, Hardecker’s article continued to describe the cross-departmental navigation necessary for the project engineer to guide his ward through development and into production. “With respect to each of his projects, he is the source of source of information for all departments and is particularly the liaison man between the customer, the manager, the engineering department, the testing department and the production department” (1926a, p. 641). Within the overall organization, the project engineer Hardecker described occupied a fairly lowly place but was required to interact directly with individuals throughout the organization and was furthermore considered an executive over the design work itself. At each stage in the engineering work, there were special forms used to document decisions and track the progress of the work.

Hardecker’s articles reproduced scheduling sheets for following tasks by personnel assigned and by project. On large projects, preparing these forms was the responsibility of a “schedule clerk” and served to keep the entire organization abreast of the work. Hardecker’s (1926b, p. 664) “form for scheduling work according to design details” is essentially a Gantt-type chart that breaks down the design of the aircraft into component assemblies—such as wings and tail—and within these to further subassemblies—rudder, elevator, and so forth. Drafting these assemblies was a team effort (Hardecker used the term “squad”), and elaborate crosschecking and approval procedures were used to ensure compatibility of both “local and foreign” drawings—that is, those produced in-house and those from subcontractors.

System Building, circa 1910

The foregoing cases make clear that the issues now associated with project management long predate “project management.” Though terms such as matrix organization, concurrent engineering, project teams, stakeholders, performance indicators, and the like would have been foreign to these individuals, similar concepts and issues were part of their working lives. What marks these accounts and helps differentiate among them is the progress of the formalization of such ideas. The question becomes “why did these people need to explain their work?”

The history of modern management is often traced to the work of Frederick Winslow Taylor, a mechanical engineer and, most famously, the author of The Principles of Scientific Management (1911). In this book and in various articles that preceded it, Taylor set out bases for the rational organization of repetitive labor processes. According to Taylor, through careful measurement of the time and movements involved in a task, the “one best way” of doing a job could be identified and taught. Through wage incentives and coaching, workers’ self-interest could be awakened to the common benefit to be achieved by doing things Taylor’s way. While Taylor was neither original nor alone in preaching the benefits of rational management (Yates, 1989; Litterer 1959), his work provides a useful marker for the general development of a formal discourse on management in the United States. It also exemplifies the emphasis in the early management literature on organizing for repetitive processes. Extended to mass production, factory systems held sway over subsequent developments in business administration and popular ideas of the means to material progress for much of the twentieth century.

W. Milnor Roberts’s reflections from the perspective of 1878 acknowledged the increase in the complexity of both business and engineering that had occurred over the course of his lifetime. Writing after 1900, Robb, McCullough, and Hardecker all sought to explain their work against as well as with the rhetoric of efficiency and system marked by Taylor’s writings. In a sense, they were all reacting against the first management fad. In light of the popular embrace of efficiency and system during the Progressive era, they were faced with explaining the limits of formal organization.

At the same time, all three needed and used the tools of systematic management. In the late nineteenth and early twentieth centuries, big construction projects—canals, bridges, and roads—were increasingly part of the coordinated making of technological systems. As these systems grew in complexity and as their making came to rely more on esoteric technical knowledge than on traditional practices, the need for formal communication on and among work sites had increased. Building a canal around 1830 could be accomplished using traditional methods. Aspects of the canal’s engineering that required innovation could be worked out incrementally in the field and the rhythm of work was still substantially dictated by cycles of agricultural employment and weather. While an early canal would eventually be a productive, revenue-generating resource, capital markets and accounting tools had not developed to the point where the cost of each month’s construction time could be calculated against future revenue. By the late nineteenth century, with the advance of factory systems, the development of formal accounting tools, and the increased pressures of capital markets, time and money had a much clearer relation to one another. Robb’s organization, Stone & Webster Company, was as much a finance house as it was an engineering organization. Though it earned a profit from engineering projects, its engineering work existed in support of its management of public utilities.

Contracting organizations and irregular production could use supervision, specification, and formal accounting in so far as economically possible. But as the individuals cited above make clear, coordination on technical projects required decentralized decision-making. Their need to balance formal and informal controls and ultimately to explain and defend a kind of work occasioned their articulations of the distinct aspects of the organization of engineering production and system building as opposed to system operating. In short, by the end of the nineteenth century a complex set of enabling and, collectively, sufficient conditions for the formalization of project management existed. The catalyst for the emergence of a professional discourse on the management of projects, though, came in the form of external pressures. Harvey Sapolsky (1972) has described the use of PERT during the Cold War as a rhetorical and representational device that served political purposes as much as it actually guided work. John Lonnquest (1996) has similarly shown the uses of “concurrent engineering” by the U.S. Air Force’s ballistic missile programs during the 1950s to be as much rhetorical flourish as genuine innovation.

This is not to cynically dismiss the importance of these or other ideas, techniques, and rhetoric that make up the discipline of project management. Rather, informed by a sociological understanding of underlying distinctions between projects and other forms of production, the episodes from the history of the project management described here emphasize the need to recognize the interplay between internal and external sources and uses of such tools.


Tracing the history of project management—by whatever name—is a matter of looking for times and places in which new industries and systems have been under construction. The codification of project management in the mid-twentieth century occurred during a period of system building in the aerospace and electronics industries and the explosion of interest in project management in the 1990s has a accompanied a period of system building in telecommunications. Similar circumstances have existed in railroad engineering in the mid and late nineteenth century, in chemical and electrical engineering at the turn of the century, and in aircraft engineering in the 1920s and 1930s. Integrating the work of scientists, engineers, technicians, skilled tradesmen, and general laborers, building a street railway or a hydroelectric plant or a chemical processing facility required coordination across fields and firms. The transferability of management techniques among these seemingly disparate worlds of work in turn reflected structural kinships between their respective labor and communication processes. That software engineers carry titles such as “solutions architect” and work toward “builds” is not coincidental. The social order of construction work and the professional culture of modern high-technology enterprise were and are closely parallel. In both, craft traditions and administration persist.

Steven Barley and Julian Orr (1998), among others, have discussed the emergence of a broad class of white-collar technical workers who labor “between craft and science” in that they do self-guided, skilled jobs to meet specifications that have been set by others. In adopting and adapting their title, I suggest that project management—though not always known by that name—has long served to both explain and insulate a kind of work, mediating between craft labor and systems of management.


Barley, Stephen, & Julian Orr, eds. (1998). Between craft and science: Technical work in U.S. settings. Ithaca: ILR Press.

Boutinet, J.P. (1996). Anthropologie du projet. Paris: Presses Universitaires de France.

Bugos, Glenn E. (1993, Fall). Programming the American Aerospace Industry, 1954–1964: The Business Structures of Technical Transactions. Business and Economic History, 22 (1), 210–222.

Dinsmore, Paul. (1999). Winning in business with enterprise project management. New York: American Management Association.

Hardecker, John F. (1926a, April 22). The functions of a project engineer in a technical organization. American Machinist 64, (16), 641–642.

Hardecker, John F. (1926b,October 21). The functions of the design division in a technical organization. American Machinist 65 (17), 663–666.

Johnson, Stephen B. (1997). Three approaches to big technology: Operations research, systems engineering, and project management. Technology and Culture 38, 891–919.

Kerzner, Harold. (1992). Project management: A systems approach to planning, scheduling, and controlling. New York: Van Nostrand Reinhold.

Litterer, Joseph. (1959). The emergence of systematic management as shown by the literature of management from 1870-1900. Ph. D. dissertation, University of Illinois.

Lonnquest, John C. (1996). The face of atlas: General bernard schriever and the development of the atlas intercontinental ballistic missile, 1953-1960. Ph.D. dissertation, Duke University.

Lundin, Rolf A., & Christopher Midler, eds. (1998). Projects as arenas for renewal and learning processes. Boston: Kluwer Academic.

McCowan, Kelly. (1999). Project management for manufacturers: The final frontier. PM Network (May).

McCullough, Ernest. (1915, February). Comments to Thompson, Sanford E. & William O. Lichtner. Construction Management. journal of the Western Society of Engineers XX, (2), 130–135.

Robb, Russell. (1910). Lectures on Organization, Boston (?): privately printed.

Roberts, W. Milnor. (1878, August). Reminiscences and experiences of early engineering operations on railroads, with especial reference to steep inclines. American Society of Civil Engineers Transactions VII, 197–215.

Sapolsky, Harvey M. (1972). The polaris system development: Bureaucratic and programmatic success in government. Cambridge: Harvard University Press.

Steiger, Thomas L., & William Form. (1991, August). The labor process in construction: Control without bureaucratic and technological means? Work and Occupations 18, (3), 251–270.

Stinchcombe, Arthur. (1959). Bureaucratic and craft administration of production: A comparative study. Administrative Science Quarterly 4, 168–187.

Stinchcombe, Arthur. (1990). Information and organizations. Berkeley: University of California Press.

The Economist. (2000, January 29). Career Evolution.

Thompson, Sanford E., & William O. Lichtner. (1915, February). Construction Management. Journal of the Western Society of Engineers, XX (2), 109–129.

Yates, JoAnne. (1989). Control through communication: The rise of system in american management. Baltimore: Johns Hopkins University Press.

This material has been reproduced with the permission of the copyright owner. Unauthorized reproduction of this material is strictly prohibited. For permission to reproduce this material, please contact PMI or any listed author.

Proceedings of PMI Research Conference 2000



Related Content