A History of Mechanical Inventions: Revised Edition

By Abbott Payson Usher

"The book is without peer in its field." — American Scientist
In this completely revised and enlarged edition of a classic work in the history of technology, a noted scholar explores the importance of technological innovation in the cultural and economic history of the West.
Following an introductory discussion of the place of technology in economic history, the author offers a penetrating historical analysis of social change. Within this context he develops a theory of invention based on Gestalt psychology and a concept of social evolution as continuous development from antiquity to the present. Emphasis is placed on the role of economic forces in the development of technology, with scientific concepts also playing an important role in bringing about change.
The latter part of the book focuses on the production and control of power in general, and in particular on a number of important operative mechanisms. Thus we read thought-provoking accounts of the technology of textile manufacture from primitive times, of water wheels and windmills, water clocks, and mechanical clocks, and the work of Leonardo da Vinci. The development of printing is carefully studied, not only for its intrinsic interest, but because of its importance for the history of science. Other topics include the production and application of power (1500–1830), machine tools and quantity production, the production and distribution of power since 1832, and the role of Asia Minor as a source of techniques which dominated the Middle Ages and the modern period as well.
Thoroughly researched and cogently reasoned, A History of Mechanical Inventions belongs in the library of anyone interested in the history of science and invention, as well as the relationship of technology to economic and social history.
"Throughout the book there is constant proof of the author's wide learning and varied intellectual interests." — The New York Times

• Language: English
• Publisher: Dover Publications
• Release date: Jul 24, 2013
• ISBN: 9780486143590

Book preview

Index

CHAPTER I

The Place of Technology in Economic History

I

Economic history is concerned with the description and the analysis of the mutual transformations taking place between human societies and their environment. The study of costs and prices is important, and the institutional structure of organized social life demands careful attention, but the basic problems of economic history lie in the field of the management of resources. We are constantly faced with the need of describing the resources effectively available to a given society over an appreciable period of time—the remote future no less than the most immediate present. The quantitative analysis of economic activities requires study of the processes and accomplishments of the system of production in physical units as well as in value units.

Economic history is therefore intimately concerned with the explicit geographic environment, and with the techniques by which resources are utilized at any given moment. The economic historian must use with care the broader results of both the geographers and the technologists. The dependence of economic history upon geography has long been recognized, but the importance of an effective understanding of technologic development is not so generally appreciated. Furthermore, little attention is commonly given to the relation between the geographic environment and the technology that makes the environment useful.

In the study of geology and geography we are prone to think of the earth comprehensively, without regard to the existing state of our knowledge of climate and resources or to our skills in using the physical environment. Such a point of view is valid for many purposes, but it is dangerous and unsound when we are concerned with the economic analysis of environmental factors. For purposes of economic and social activity, the geographic environment is not the totality of physical features, but only that part of the complex which we can conceivably use, immediately or ultimately. This effective geographic environment is determined by our skills in using it; it is, therefore, related to the development of technology. The environment is enlarged by new knowledge and new skills. The distinctive feature of human evolution lies in this fact. Human societies not only select an environment, they make their environment. The processes by which man makes himself include those procedures by which men transform their environment. Human evolution is doubly dynamic; man and the geographic environment react upon each other, and both terms are transformed.

Broadly conceived, technology is an important part of the central core in the evolutionary process. It is an essential aspect of the accumulation of knowledge and the development of skills. It does not exhaust the field of the development of the mind, but it is a characteristic segment of the whole. If we may presume that some common principle underlies all our mental activity, technology, and most particularly the mechanical field, may possess a peculiar importance. The processes of innovation can be studied more conveniently in the mechanical field than in most of the conceptual fields, because mechanical apparatus can be traced more accurately and more completely than religious, ethical, and philosophic concepts. In its own right, and as an aspect of the general process of innovation, technology has powerful claims upon our attention.

The central importance of technology will be more accurately perceived if we consider in broad outlines the relation of technical change to the geographic environment. The physical features of the environment may be divided into three general groups: the soil-climate complex, mineral resources, topographic structure.

The classification and description of climates has been developed on several principles, all of them significant and useful for particular purposes, but no single system affords direct and conclusive evidence of the economic importance of the different climates.¹ The simpler systems of classification rest upon the analysis of rainfall and temperature data. It is essential to recognize seasonal variations and these adjustments present many difficulties when it becomes essential to determine the boundaries of climatic zones with accuracy. For some purposes averages are adequate; for other purposes full data on the range of variation are indispensable.

Thornthwaite has developed a refinement in the analysis of the temperature and rainfall data by computing the amount of evaporation that may be expected under the various temperature conditions. A coefficient of effective precipitation is developed that presents a more delicate measure of the boundaries of climates. The redefinitions are especially important in the distinctions drawn between the arid and the subhumid climates, and a more accurate picture is given of the effective differences between the warmer and the cooler climates.

An entirely different approach is afforded by the analysis of the natural vegetation of different regions.² It is presumed that the actual vegetation presents a more accurate picture of the combined effect of all factors than any computation based on temperature and rainfall records. It has also been presumed that natural vegetation presents a better measure of the economic significance of the different climates. These studies of vegetation afford both some measure of the influence of differences in soils and some index of soil characteristics. We now know that soils are a complex product of the whole array of factors; parent materials, rainfall, temperature, and the character of the vegetative cover.

Natural vegetation brings us closest to the problems involved in the study of the development of agricultural activities, as the early forms of agriculture are closely related to natural vegetation. It is necessary to distinguish at least three general types of activity even in early cultures; dry-land agriculture, wet-land agriculture, and the purely pastoral activities of the nomads. Some considerable portion of the areas suitable for these different activities is probably determinable within reasonable limits of accuracy, but there are appreciable zones at all boundaries that allow freedom of choice. In these zones both political and technical factors may exert powerful influences, so that no determinate statement of resources can be made unless we have knowledge of social and economic conditions.

It is presumed that the early agriculture in northwest Europe was largely confined to the loessal grasslands, and that the clearing of the land for the plough did not begin to spread into the forested areas until the fifth and sixth centuries of the Christian era. Medieval and early modern agriculture was occupied with this process of picking over the forested land. Seventeenth-century colonists in North America had become so accustomed to clearing forest land that many supposed that the best arable land had a natural cover of hardwoods.

Since the seventeenth century, the outstanding developments in land use involve two types of change. General farming has been supplanted in many areas by the culture of special commercial crops in areas peculiarly suited to their use. But an even more important change has grown out of the new relation between cereal culture and stock raising. In the older systems of agriculture the products of the arable land were used almost exclusively for human food and alcoholic beverages. The stock derived some food from the mature straw of the cereal crops and from grazing on the arable fields after the harvest, but the effective forage derived from these sources was small. Livestock were primarily dependent upon the unimproved land—grassland, mountain heath, or forests. The most substantial pastoral cultures were nomadic: pure nomads moved their herds about in areas without any permanent settlement; the seminomads maintained a culture independent of the settled farmers, moving herds on regular routes between winter and summer pastures.

The development of the cultivated grasses and legumes resulted in a profound change in the balance between pasture and tillage, and ultimately resulted in great increases in agricultural productivity. Agriculture in North America presented special patterns of land use associated with the position of maize, potatoes, tobacco, and cotton.

Monsoonal Asia presents special problems which are not essential to the immediate discussion, but the spectacular consequences of the establishment of the sugar culture in Brazil afford a striking illustration of the relation of technique to resources. Though not indigenous to the New World, the sugar cane proved to be well adapted to a wide array of the soil-climate types found in South America and the Caribbean. Land ceased for a long period to be a limiting factor in the spread of the sugar culture.

The purely objective descriptions of climate or of the soil-climate complex afford no determinate definition of agricultural resources for two reasons: first, boundaries of the various patterns of land use in agriculture do not coincide with the boundaries of the objective classifications of climate; second, the use of important soil types depends upon accessibility to markets, either because the best product is too specialized to enter largely into general farming or because there is too much land of the given type to be fully used. Areas suitable for sugar and tobacco fall within the first type. In major centers of production, the soils of these areas could be used for general farming, dominated by the production of food for local consumption, only on a very restricted scale. Highly specialized small-grain areas in the drier margins of cultivation illustrate the dependence on economic and technical conditions. The wheat fields of Southern Russia and the wheat areas in North Central United States and Canada could not be conveniently or fully utilized for general farming, and much of the area could scarcely be utilized at all except as a very highly specialized region dependent upon cheap heavy-duty transport of massive raw materials. Agricultural resources are, therefore, determinate only in terms of some defined technology of production and transportation. The techniques existing at any particular time can be projected into the future for an appreciable period, but such projections are subject to the qualification that the estimate represents our present point of view and our present anticipations of applications of known techniques.

The error of the geographic determinists lay in their failure to recognize these limitations in the description and analysis of physical resources. They were too ready to view all the past and all the future in terms of existing knowledge. In this way they adopted a misleading and inaccurate concept of the geographic environment. This environment always belongs to some specified social group, however primitive or however sophisticated. We cannot identify our present position with the whole course of geologic time, entirely apart from the fact that present-day conditions have really occupied only a very small fraction of the known geologic period.

In respect of minerals, similar difficulties arise, though in slightly different forms. Let us assume that a survey of mineral resources was being made about 1820, for the world at large or for the United States and Europe; would it be essential to include the known resources of petroleum, scanty as they were? A few outcrops were known in the United States, the oil shales of Scotland were known, and in the Caspian a limited commercial use was made of the crude oils obtained from hand-dug wells. If the basis of the survey was defined as an estimate of reserves of commercially useful minerals, the resources of the United States and of Scotland would not have been included, for at that time there was no known technique for the use of crudes in either country, and no knowledge of the possible processes of distillation. Lack of commercial interest resulted also in an almost complete neglect of the various evidences of the existence of petroleum.

The actual difficulties of estimating petroleum reserves present a different and well-nigh unique problem. The mobility of petroleum and its concentration in a variety of traps among the impervious strata make its measurement underground impossible. We are thrown back upon estimates of the life of wells and of fields, and upon the probable content of the heavier deposits of sedimentary rocks. These difficulties of estimate obscure the importance of the development of new techniques in drilling and in the detection of oil in formations that give no surface indication of its presence. The great increases in the production of petroleum have been made possible by great increases in the depth of drilling. In the early years, Pennsylvania crudes were obtained at depths that rarely exceeded 300 feet; today large quantities are obtained from wells of more than 8000 feet, and producing wells have been brought in at 15,000 feet. The discovery of new fields has attracted wide attention, but new techniques of recovery are scarcely less important.

The deposits of fixed minerals commonly admit of description in terms of the depth of the ores, the width of seams, the approximate concentration of the minerals that are diffused in a matrix. Deposits of gold, silver, tin, copper, iron, coal, and most other important minerals can be described completely in terms of the occurrence of the mineral without explicit reference to specific techniques of exploitation. The interpretation of the record, however, requires consideration of techniques and costs of extraction. Determinate quantitative estimates of resources involve some positive definition of the technical conditions presumed to control the production of the mineral. Estimates usually distinguish between actual and potential reserves, the latter being reserves that cannot be worked by present techniques and at present costs. In practice, it is difficult to secure uniform definitions of the boundary between actual and potential reserves, so that totals for the world or for major continental areas require careful interpretation. Estimates of mineral resources are, therefore, based on the principle that the techniques of exploitation are defined.

It might be supposed that topographic features are purely objective, and that their description and analysis do not involve any consideration of technology. Subject to continuing processes of geologic change, the face of the landscape must inevitably seem to be one of the most objective elements in our environment. But as soon as man comes upon the scene, topographic features enter into human activities and in many respects are dominated by them. The influence of human settlement upon the environment is shown by the changing importance of urban sites.³ Sites derive their significance from the combination of local and regional functions. The site is useful both as residence for its permanent population and as a focus of regional activity. The town site may serve as a fortress to which the population of a considerable area may retire if threatened by force. It may be a center for trade, for industrial production, for civil or ecclesiastical administration, for cultural activities, religious or secular, for recreation. The use of a site is thus a definite outcome of an array of economic and social activities that have become associated with it, so that its life and development are peculiarly contingent upon human interests.

Although the defense functions of a site might seem to place a premium on inaccessibility, the important defense features are immediate and local, and are subordinate to other features. Defensive sites are really sites that present special opportunities for local defense, despite the fact that they command major land or water routes. Sites derive their basic values from their accessibilities. They are focal points in the network of land routes, and transfer points between land and water transport. Mackinder described them as nodes in the transport system. As a basis for units of settlement, the site becomes an organic part of the total social structure. The size of the individual units of settlement is a function of the total population of the region. The various units of settlement disclose a systematic gradation in size that exhibits the characteristics of a harmonic series. The structure of settlement in any region is an organic whole that is the result of dynamic adaptation to the environment.⁴ The system of transport is organized with reference to the physical features of the area, but it is also a function of the technology of the particular society.

The significance of the changes in the techniques of transport can be most accurately appreciated if we consider the effective capacities of the different forms of transport. On the basis of experience in Africa and other pioneer areas, Holmstrom has estimated the capacity of transport by porters and animals.⁵ Human porters, if fully organized for systematic transport, can move 1750 tons of goods per route mile per year per man. Light pack animals were rated by Holmstrom at 1590 tons per route mile per year. Heavy pack animals were rated at 3600 tons. A bullock wagon can move 8640 tons per route mile per year. Horse-drawn wagons have an effective capacity of 15,000 tons per route mile per year. If we make a similar estimate for inland waterways, the importance of water transport is vividly revealed. Upon a natural waterway with one towing path, assuming a speed of 2.5 miles per hour and six barges per mile, it would be possible in a 12.5-ton barge to move 2250 tons of freight per route mile each twelve-hour day. If navigation were possible for 220 days in a year, 495,000 tons of freight could be moved per route mile per year. Deep-water transport is, of course, capable of greater volume of service than inland waterways, because larger units can be used and the number of units is not as rigidly limited. It would be possible to interpret extant materials for antiquity and the Middle Ages with sufficient accuracy to establish the approximate capacities of the systems of transport that have dominated social life throughout the historical period. But it is not essential to have all the details. It is quite clear that heavy-duty transport was restricted to deep and inland waterways until the establishment of the steam railroad. As the railroad networks developed in North America and in Europe, life was profoundly transformed. The change was most sharply emphasized by the spectacular growth of the great metropolitan centers, but it was not confined to the primary sites. The balance between urban and rural life was almost completely reversed and the influence of the new forms of economic structure were manifest throughout the entire pattern of settlements.

Changes in the industrial techniques should not be ignored, but they have been complementary. Urban units develop in association with massive resources of fuel and iron, but these developments were in part conditioned by the development of a comprehensive system of heavy-duty transport.

The dependence of site values upon technologies of transportation is illustrated vividly by the changes in the relative importance of Philadelphia and New York in the eighteenth and nineteenth centuries. For purposes of comparison it is important to use statistics for Philadelphia County, since the municipal area of Philadelphia was small and had already ceased to represent the entire urban settlement. On this basis, Philadelphia was larger than New York until 1810 and about equal to New York in 1820; after 1830 the primacy of New York became steadily greater.

The early development of New York was influenced by the restricted area of good land in its immediate vicinity. The settlements served by New York were confined to the immediate neighborhood and to the Hudson and lower Mohawk valleys. The Hudson-Mohawk route was of limited value because settlement was too thin to require any considerable volume of transport. With increased densities of settlement and the completion of the Erie Canal and the branches of the state canal system, the entire picture was changed. The development of the Philadelphia market was hampered by the Alleghenies and its growth slowed down just as New York began to develop more rapidly.

With the development of the railroad network the position of New York became more commanding, because facilities for movement by rail narrowed the issue to a comparison of port facilities, and on this basis the ports of Philadelphia and Baltimore were unable to compete. Despite the efforts of the Pennsylvania Railroad and the Baltimore and Ohio Railroad to build up their own ports, both were ultimately forced to recognize New York as their primary terminus. The traffic of the trunk-line territory was concentrated in New York for export, and later the traffic of the entire Pacific coast. Improvements in transportation made the properties of the site of New York progressively more important. Even a difference in the timing of the development of rail transport would have altered the history of these cities. There are, therefore, many grounds for insisting upon the interdependence of environment and technology. The precise significance of environmental factors is dependent upon the techniques available to a given society for the use of its resources.

The relation between technology and resources affords an outstanding illustration of what William James called soft determinism. At any given moment resources are determinate and limited in amount. The limiting volume of resources is always present in the two categories of immediate and ultimate resources, or, in the usual terms, as actual and potential resources. But these limits are not absolutes, as is presumed by the geographic determinists. The limitations of resources are relative to the position of our knowledge and of our technique. The limits of both actual resources and potential resources recede as we advance, at rates that are proportionate to the advance in our knowledge.

These concepts are so commonplace and so persistently recognized in economic literature that it is not easy to understand why it has been so difficult to escape from static and deterministic concepts. The explanation is probably to be found in the time dimensions of technical changes, and in the difficulties of giving effective mathematical expression to this interdependence between resources and technique. The full compass of technical change is spread over longer periods than it is convenient to treat in economic or geographic analysis. The effects that are readily distinguishable seem to be sharp and sudden in their incidence, so that we are easily led to suppose that changes in technology are rare and exceptional occurrences, which can legitimately be treated as unusual events, exceptions to the prevailing conditions of rigid limitation. Acknowledgment of the effects of improvements in the arts would seem to be adequate, even if confined to the qualifying clause, unless there are improvements in the arts. Such attitudes were fostered by inaccurate knowledge of the history of pure and applied science.

It is important not to presume a continuous development of technology at a constant rate, but it is important, also, to recognize that the process of social evolution consists in part in the cumulative development of science and technology. We need both a general understanding of the process or processes, and, when records make it possible, a documented account of the history of particular periods and particular achievements.

II

Effective historical analysis is dependent upon some mathematical and statistical technique capable of expressing these interdependent relations both in purely theoretical form and in statistical analyses of specific economic phenomena. Attempts to apply mathematical techniques to the analysis of phenomena of growth begin significantly in 1846 with the efforts of Ver Hulst to give an adequate formulation to the simpler generalizations of Malthus about the growth of population. Ver Hulst found an elegant solution to this problem in the autocatalytic or logistic curve.⁶ The formula is designed to describe the conditions of the growth of a population that has not fully occupied its area, nor made full use of the resources open to it with the techniques used in production. It shows that the limitations of resources are constantly present, so that growth takes place at a continuously decreasing rate.

Little interest was shown in this analysis for many years, until it was applied to population problems by Raymond Pearl in a well-known series of articles and books.⁷ Pearl was unwilling to recognize the gap between the theoretical formulation of the problem and the empirical phenomena of population growth in the eighteenth and nineteenth centuries. He applied a specifically static formula to one of the most dynamic periods of accelerated growth in population of which we have any knowledge. The mathematical and statistical features of Pearl’s position have been extensively criticized, so that no discussion is now necessary. His thesis concerns us here only because it shows how easily a dangerous concept of resources can be accepted by a scholar of high attainments. The entire conceptional framework of Pearl’s logistic analysis is inconsistent with any concept of evolutionary process or with any adequate interpretation of the empirical data now available to us.

Mathematical analyses of primary economic phenomena have latterly been associated with trend curves and business-cycle studies. These approaches to the problem have been essentially empirical. The problem assumes the form of fitting a curve to the data obtained by various processes of smoothing. The literature has been surveyed with care by Frickey and by Kuznets.⁸ They find no determinate solution in the empirical data now available. No single curve can be fitted with a sufficient degree of accuracy to justify any belief in its unique significance. There is sufficient conformity between the computed curves and the smoothed data to warrant the conclusion that the phenomena disclose some kind of orderly progression, but we are not now in position to assign primary significance to any particular formula. Frickey is convinced that further advance can be achieved only by a more rigorous and general theoretical analysis. There are, therefore, many reasons for changing the method of approach. Emphasis may best be shifted from the details of curve fitting to the consideration of the significance of the characteristics of the various curves and their relevance to the phenomena we are seeking to analyze.

From this point of view the general framework of the normal curve offers a very suggestive approach to the problem of expressing the functional relations between techniques of production and total resources. The use of this curve was first suggested in 1916 by R. A. Lehfeldt, but without extensive discussion. In the recent literature on trend analysis, it seems to command little interest or enthusiasm. As a basis for curve fitting over relatively short periods of time it offers no special advantage and no clear superiority in the fit secured to the data. The computations involved are more laborious, and as a consequence other curves have been preferred. These severely practical disadvantages are less important if the emphasis is shifted to the general properties of the curve without regard to the details of computed curve fitting.

The curve suggested by Lehfeldt is the integral of the normal curve, in which the changes are symmetrical to the logarithms of the items, rather than to the items themselves. Lehfeldt describes the formula as follows:

"Let q be the quantity whose changes in time t are to be studied. It is not to be expected that the changes of q itself should be symmetrical in time, for all the changes observed in the latter half of the period of change refer to values of q larger—possibly many times larger—than in the earlier half. But log q may very possibly undergo symmetrical changes, so we will assume that it is a ‘normal error function’ of the time, i.e.,

log q = log q0 + kF(t/T),

where q0 is the value of q at a certain moment (the ‘epoch’), t is the time in years before or after the epoch [the point of inflection], T is a constant period, and

and k is a constant."⁹

Such a curve represents a process of change at rates that increase progressively up to the point of inflection and decrease progressively after the point of inflection is passed. It is, therefore, a description of growth taking place under the combined influence of factors that produce acceleration and of limiting factors that prevent any continuous or constant acceleration. The general framework of the curve is certainly expressive of some such organic relation between technology and resources as we can establish by general empirical analysis. It remains to discover whether it is possible to establish any reasonable presumption that the curve is relevant to the interpretation of the statistical data on economic growth.

Lehfeldt gave illustrations in his article of applications of the formula to the population of England and Wales between 1660 and 1960 and to the development of the external trade of England. Applications of this technique of analysis to the records of population growth are especially important since the data reflect in the broadest manner the total array of technical and resource factors. Figure 1 exhibits the population data for England and Wales plotted on a semilogarithmic grid. It is possible to use an arithmetic grid for graphic presentation, since it is a matter of indifference whether the items themselves are plotted on a logarithmic scale or the logarithms of the items are plotted on an arithmetic scale. For purposes of computation the logarithms of the items must be used. A smooth curve has been drawn through the points plotted from Lehfeldt’s data. Other estimates have been plotted for the precensus period, and later census figures have been added to Lehfeldt’s series, since he was writing in 1916.¹⁰ In so far as the curve exhibits accurately the formula for the probability integral, the curve is symmetric with reference to the upper and lower asymptotes for points at equal time intervals from the point of inflection. Thus, the points plotted for 1900 and 1780 should be symmetrically related to the framework of the curve. The linear distance of the item for 1780 from the lower asymptote should be equal to the linear distance of the item for 1900 from the upper asymptote, and conversely. The pure curve should approach the limits without at any time actually reaching them. This condition is never realized in actual social phenomena. The lower asymptote is in all cases actually passed, and after primary technical change the upper asymptote of one period of development becomes the lower asymptote of a new phase of growth. The framework of the curve gives objective definition to the boundary lines between periods of growth. The lower asymptote represents the total resources of the society immediately before a great advance in the technology of resource use. If quantitative expression is to be given to resources, both actual and potential resources should be included. The upper asymptote represents the totality of resources after a great technical advance. In actual life the passing of the limits of what was at one time an upper asymptote is accomplished as the result of massive innovation. In England at the close of the seventeenth century the new technology centered around increased use of coal in industry, culminating in the development of techniques for the use of coal as a source of power and as a metallurgic fuel. These changes gave an entirely new valuation to the coal resources of England and Wales.

Fig. 1. Population of England and Wales, 1650–1960, plotted on a semilogarithmic grid. The framework of the integral of the normal curve is shown in broken lines. Rickman, +; Griffith, ▲; Lehfeldt, ○; Census, ●.

Fig. 2. Population of England and Wales, 1650–1960, plotted on a probability grid. The percentages of the vertical scale are derived from the division of logarithms of the various figures for population by the difference between the logarithm of the upper asymptote and the logarithm of the lower asymptote. Rickman, +; Griffith, ▲; Lehfeldt, O; Census, ●.

As the upper asymptote is approached, the limits of resources become progressively explicit, but since it is unsound to extrapolate, we should never presume that the future is rigidly determinate. Basic changes may occur at any time which will affect the primary functions underlying growth. In extreme cases the growth function may assume a negative value. The best illustration of such an event is the history of population in Ireland after the onset of the potato blight in 1846. Population declined, in the general pattern of a reversed normal curve. The features of the probability integral thus express in highly generalized form the relations between technology and resources that may be inferred from the crude empirical data without any formal mathematical analysis. The curve does not express a law of population growth, but it does show that the interdependence between technology and resources results in orderly growth over long periods of time. These relations must be dominant factors in the secular trend underlying primary movements in resource utilization.

The features of the probability integral are, therefore, suggestive, but it is of course unwise to make extensive use of this technique of analysis unless the application to empirical data exhibits enough conformity to the formula to justify generalization. Study of the relative value of this curve has been obstructed by the lack of long series of statistical records of population or production. The distinctive features of various curves do not appear clearly unless very extensive portions of the curve can be studied. Many of the time series now available are too short to admit of a determinate mathematical solution. Lehfeldt assumed that no adequate empirical verification could be made of the probability integral because of the deficiencies of statistical material, so we may presume that his lack of interest in further work on the problem was due to this feeling that no serious study could be made. It is probably true that we have not now at hand statistical materials that would suffice for a wholly satisfactory demonstration of the significance of the probability integral in this connection, but more material is available than is commonly recognized, and graphic techniques of analysis are available that make it possible to use material that would not be adequate for the computations required by comprehensive algebraic analysis.

The fit of the curve to the data can be tested by expressing the items as percentages of the difference between the upper and lower asymptotes. If these percentages are plotted on a special grid that expresses the probability function, a true curve becomes a straight line. Lehfeldt’s data have been plotted on a probability grid, with other data on population of England and Wales for the period (Fig. 2). There are many features of interest in the results. Lehfeldt’s figures do not conform precisely to the requirements of a true curve. Within the census period, the figures for 1900 and 1810 are considerably off the line. The figures for 1660 and 1690 are very much in error, though they are lower than any other significant estimates for that period. The census items for 1931 and 1947 are considerably above the line, though they represent events subsequent to the original calculation. All estimates for the period prior to 1720 are substantially higher than the figures implicit in the calculated curve, and, curiously enough, the recent growth of population is in excess of the expectations suggested by the curve. In terms of the mathematics of curve fitting the deviations from the curve are considerable and of such a nature that the data do not constitute an empirical verification of the curve as an unqualified description of population growth in this period. One must presume that factors are involved that are not consistent with the postulates of the curve. The simplest explanation of the nonconformities at the beginning and end of the period is to be found in the extent of technical improvements at these times. The actual beginning of new advances in technology was in both instances more gradual and more considerable than the general curve would suggest. The implications of the curve are likely to be unrepresentative in the early and late phases of economic development.

It is not easy to determine the significance of an analysis that involves nonconformities of these types. It is necessary, however, to recognize that this series of data covers a total period of 300 years, and that conformities are certainly significant for a period of fully 200 years (1720–1930). The curve suggests a degree of orderly progression in growth that is not commonly suspected, and the underlying implications of the curve are more suggestive than the relations implicit in other curves. The probability integral affords an explicit account of the characteristics of a functional relation between resources and technique. Technical advance modifies the limitations imposed by the external environment, but the effective limits of resources can never be neglected. Truth lies between the extravagant pessimism of extreme geographic determinists and the equally extravagant optimism of those who believe that applied science overcomes the niggardliness of nature that worried Malthus and the Classical Economists.

CHAPTER II

Historical Analysis of Social Change

I

The conventions of historical writing have always stressed presentation in narrative form. Criticism plays its part in the study of documents, in the search for explicit evidence of the chronology of events, and in the elimination of conscious or unconscious misinterpretation of motives and acts. The implied antithesis between scientific analysis and historical narrative has, however, constantly obscured the importance of analytic procedures in history, though for more than a century there has been a steady development of techniques of analysis that are closely comparable with the techniques of the sciences in all their operational details. Despite some minor differences in the character of the problems involved, we must now assume that the logical and epistemological problems faced by the historian are not essentially different from the problems of the scientist. If we are concerned with the development of an institution like Parliament in England, significant understanding of the development requires accurate analysis of many changes in form and function. Changes in function require new forms; new forms foster further changes in function. A large measure of identity was preserved over a very long period, despite the magnitudes of changes in function. The history of Parliament can doubtless be presented most effectively and most compactly in narrative form, but the narrative must rest upon a great deal of careful analytic work. The problems of growth and change involved in the history of Parliament are neither more nor less complex than the problems presented by paleontology or genetics. The materials are different; the problems are the same.

In other fields of history, analytic techniques are equally important. If we have to do with the history of thought in any field—religious, philosophic, or scientific—much analysis must precede any expression in narrative form, and in many instances direct presentation of the analytic work may be preferable to a form of presentation in which the critical discussion is omitted or obscured. The problems of economic history resemble the problems of biology even more closely, because the relations with the physical environment are so closely involved. The development of human societies requires no less careful study of ecology than is necessary for the understanding of the growth of populations of the various plants and animals.

II

Such procedures of analysis can be developed most adequately upon an explicitly empirical basis. Transcendentalism is not wholly consistent with such an approach to historical process, because the implications of substantial analysis suggest a more pervasive process of change than is provided by the great-man theory of history. More variable factors call for attention than are recognized by the highly idealistic types of interpretation characteristic of the transcendentalists. The analytic procedures that have developed in many systems of sociology and anthropology are so completely pervaded by concepts of mechanism and determinism that they cannot be used for explicit historical analysis without some modification.

Operational techniques of historical analysis have much to gain from the acceptance of the primary postulates of a consistently empirical philosophy. It is true that its problems diverge too widely from the problems of the natural and biological sciences to warrant direct borrowing of techniques or reasoning from analogy, and the historical material is too difficult and too complex to be considered as an objective verification of the empirical philosophy. But it is now within the limits of effective achievement to demonstrate modes of analysis by which the empirical point of view can be applied to historical materials. It should be possible to show that empirical concepts do not break down when applied to problems of social evolution and social change.

The analysis of the processes of social evolution in their entirety requires study of processes of invention and processes of diffusion by imitation. The analysis of cultural diffusion is no less part of the historical problem than the analysis of change. The present purpose is to analyze the processes of change with explicit reference to the place of individual effort in the general social process; but before we undertake a systematic study of invention, there are a number of general features of the historical process that should be considered.

Early concepts of social evolution were expressed as sequences of stages that were presumed to describe the entire social structure. The structures of each particular period were represented as a development out of the totality of the preceding period. The totality of the present was derived from the totality of the past. Many alternative concepts of stages were put forward in the middle and late nineteenth century, but the details are unimportant if the sequences were expressed in this linear framework as a succession of comprehensive social entities. Adequate historical analysis requires concentration of attention on particular sequences of events. To a pluralist, such a procedure would follow as a matter of general principle. To an empiricist, concerned primarily with an operational point of view, it is essential not to assume the existence of relations that have not been established by analysis. Whatever may be the scope of the relations that may be found to exist, we cannot properly begin by presuming the existence of any particular classes of relations. We must begin with the smallest possible units, and proceed from them to larger systems of relation. The principle of historical continuity must be expressed very concretely. We ought not to say that the present is derived from the past and the future from the present. The proposition must be formulated in much more specific terms: every event has its past. The principle of historical continuity does not warrant any presumption about the relations among events occurring at the same time. This assumption is very frequently made, but it will be readily seen that it is not warranted.

Distance alone leaves many culture groups or societies sufficiently isolated to prevent any significant relations among them in many phases of social life and action. The scope of cultural contacts has been increasing, so that we might justly say that the process of social evolution consists in the building up of larger and larger systems of related events, and larger organic units for social action. But if this is indeed the essence of the process, it is especially important not to assume that such intimate relations existed in the beginning, and that they continue as a comprehensive system throughout the entire course of social evolution.

Even within a closely organized social group many gradations of relatedness may be exhibited by the various systems of events that we find. The possibilities of variations in the relations among events will be appreciated more vividly if it is explicitly recognized that we must consider not only accomplished action, but also the behavior patterns that lead to action or obstruct action. The formation and development of behavior patterns in the different fields of social action should be treated as distinctive systems of events whose relations to each other and to activity is to be determined.

Max Weber raised the very significant problem of the relation of religious thought to economic activity, maintaining the thesis that Catholicism provided such an unfavorable background for certain kinds of economic activity that the origin, or at least the development, of capitalism was inhibited in the Catholic countries. No issue could demonstrate more precisely the meaning of the pluralistic approach to historical analysis. The concept of ideal culture types presumes that the interrelations among different fields of activity must of necessity correspond to the needs of internal consistency of the ideas involved implicitly or explicitly in the behavior patterns. Empirical analysis of the events themselves shows that we cannot assume consistency of relations.

The religious beliefs of the Hindus have certainly obstructed the economic development of the Hindu community over a long period in the past. Social mobility was reduced far below any level consistent with desirable changes in economic activity. Food habits were profoundly affected by the veneration of the cow, and by the attitude toward animal life. Business activity was certainly inhibited by ascetic ideals and by religious and social discrimination against trade. Specific religious beliefs undoubtedly led to less successful use of resources than would be possible under the influence of other patterns of religious thought and observance. The Hindu culture thus contained elements of divergence that were inconsistent with the best interests of the society as a whole, though the painful material consequences were obscured by the ascetic denial of material values. Criticism of the cultural system is possible once we recognize the desirability of balance among the various value systems that may conflict. But, of course, even such a modest basis for criticism rests upon a presumption that the social complex involves many values and value systems, no one of which can claim precedence or preponderant importance.

Religious belief in western Europe certainly did not exert a preponderant influence comparable to the influence of religion in the Hindu culture. However we phrase a definition of capitalism, the outstanding features of capitalistic society are to be found in the banking system; in the instruments of title, nonnegotiable and negotiable; in the developed accounting system, with double-entry bookkeeping; and in the forms of association in the use of capital, whether partnerships or corporations. The research work of the last thirty years has shown that all these basic institutions were developed prior to the Reformation in centers of powerful ecclesiastical influence. Neither the Church as an organization nor the religious beliefs as dominating patterns of behavior inhibited the emergence of all the essential institutions of modern capitalism.

Analysis of the conditions of development of northern and Mediterranean Europe after 1500 does not provide such a sharply defined conclusion. The simpler explanation of differences in the rate and character of economic development is to be found in the geographic conditions that fostered the growth of the maritime industries and the textile specialties of the northern countries.

Pluralism at least affords a basis for operational analysis with the least possible intrusion of speculative philosophy. Monistic idealisms of any type are likely to obscure important problems of analysis. If one is willing to accept recent developments of empirical philosophy and of symbolic logic, it is possible to present a systematic and integrated interpretation of the philosophy of nature and of man’s place in nature. But the judgment of these views is properly a subject for philosophic discussion. As historians we may legitimately confine our attention to the operational problems of analysis.

III

The structure of events in time reveals discontinuities of two distinct types. There are discontinuities between different systems of events that persist over long periods of time and, in many instances, without any prospect of ultimate synthesis. There are other discontinuities that may be overcome, through some act of synthesis. The establishment of new organic relations among ideas, or among material agents, or in patterns of behavior is the essence of all invention and innovation. Analysis of the process of cumulative synthesis will occupy us in the following chapter. For the present it is desirable to describe in more detail the form and character of the discontinuities that persist.

A comprehensive classification of the kinds of material to be analyzed would be confusing. It should be enough for the present purpose to describe the primary classes. The historian is concerned with thought and action at three distinct levels: abstract ideas and concepts; patterns of behavior expressed as habits or policies; explicit action. The elements of social change appear first as new ideas and concepts. As the new ideas are diffused and assimilated they affect patterns of behavior and political and social action. Analysis of change in social structures and policies, therefore, requires study of underlying thoughts. To understand events in their common-sense meaning we must understand the conceptual basis of action. So we are led back to the intellectual processes themselves, and as we become more sophisticated we recognize that all thought is a kind of action. To make value judgments is to act, whatever level of abstraction may be the field of judgment.

Fig. 3. Primary classes of systems of events: 1–4, concepts and practices associated with various systems of polytheism; 5, 6 archaic features of the Roman civil law; 7–9, basic concepts of praetorian law; 10, 11, basic concepts of the Euclidean geometry; 12, 13, primary features of the geocentric concept of the universe; 14–16, concepts and practices in various monotheistic religions; 17, heliocentric concept of the planetary system; 18, atomistic concepts in science; 19, the invention and diffusion of water wheels.

There are many discontinuities among the events in different fields of activity. Philosophy is frequently in sharp conflict with religious beliefs and practices. Technology and organized scientific work are commonly somewhat set apart from aesthetic, religious, and ethical fields of action. The most fundamental discontinuities are to be found among systems of events that are related differently to past, present, and future. In the total array of systems of events, as we find them at any given moment, many systems have persisted from a remote past, and have lost all significant contact with current patterns of behavior. At best, they are conventional acts that persist without having any present meaning. At worst, they obstruct desirable forms of action. Hunting privileges, for instance, were among the most useless and most irritating survivals of feudalism. The duty of protecting the peasantry and their crops from wild animals degenerated into a right to preserve game and to hunt over the peasant’s crops, to the detriment of every vital interest of the peasant.

At any given time, a large portion of the total moving array of events is relevant to current modes of action, integrated more or less closely into a unified social system. Outside of all this material are to be found the thought and actions of men who are ahead of their time. Much of this activity may take place in fields of abstract thought, but scientific experimentation, technologic inventions, and explicit formulations of social policy cannot be excluded. The work of Galileo, Newton, and Darwin can be classed as sufficiently novel to stand outside the current of contemporary thought and action.

These relations may be symbolized in a diagram (Fig. 3). The obsolete or obsolescent systems may be represented by dashes; currently important systems, by continuous lines; nascent systems of events, by light dashes. A specific period of six hundred years has been selected in order to avoid bare abstraction, but the lines in the diagram represent only a small fraction of the total material relevant to the history of the period. In the third century B.C. the polytheistic religions were palpably obsolescent over much of the eastern Mediterranean, despite the continuance of official public worship in the temples. Some concepts and procedures of the Roman civil law were of subordinate importance as early as the third century B.C. and were undoubtedly obsolescent in the second century. The Praetorian Edict, however, was of great current importance and rapidly became the predominant element in Roman jurisprudence. Full analysis would properly require explicit discussion of each group of underlying concepts; but this large array of material can be represented schematically by the three lines shown in the diagram, which emphasize the necessity of dealing with individual concepts. The Euclidian geometry, too, should be conceived as a system of distinguishable concepts, and not as an indivisible whole. The scientific achievement of the period may

Reviews for A History of Mechanical Inventions

[bjosefsendk · Rating: 5 out of 5 stars]

very thorough and interesting from start to end. intereting perspective on our history thrpugh technology