Leonard M. Dudley †

Space, Time, Number:

Harold A. Innis As Evolutionary Theorist

Abstract

What causes economic change?  Traditionally, economists have answered that the explanation lies in exogenous shocks to technology, factor stocks, or preferences.  In the last half-decade of his career, the Canadian economic historian Harold Innis (1894-1952) proposed an alternative approach - a theory of endogenous change in communications technology.  He argued that the principal developments in western social history could be ex­plained by a process of alternation between media biased towards conservation of information over time and those biased towards transmission over distance. This paper demonstrates the close parallels between the concepts used by Innis and contemporary theories of social evolution. It also indicates the importance for future research of his vision of communications media as the most fundamental of enabling technologies.

† Université de Montréal

Harold A. Innis Memorial Lecture presented at the Canadian Economics Association meetings, Université du Québec a Montréal, Montreal, 2 June 1995.  The author thanks Pierre Fortin, Michael Huberman, Bentley MacLeod, and Francois Vaillancourt for their helpful comments and encouragement.

754

I. Introduction

What do Innocent III (1160-1216) and William Henry Gates III (1955- ) have in common other than the same roman numeral after their names?  At first glance, very little: the first was a mediaeval pope; the second is chief executive officer of a contemporary computer software firm.  However, there are striking parallels in their careers.  Both dropped out of college as young men and went on to become the leaders of powerful organizations specialized in information processing.  Both did so by exploiting new technologies that reduced the cost of storing information for elite groups within their societies.  And the contributions of both point to serious limitations in neoclassical general equilibrium theory.

The problem is to interpret economic change.  Traditionally, economists have explained changes in the price system by exogenous shocks, either to preferences, to factor endowments, or to technology.  Given such a shock, neoclassical theory is able to predict its consequences for quantities and prices.  Yet if examined closely, most of such shocks turn out to be new ideas, whether new fashions and customs, new attitudes to working and saving, or new ways of combining production factors.  Unless an economic theory provides an explanation for how ideas originate, how they are diffused, and how they are selected, it is incomplete.  In this sense, the neoclassical heritage is unable to cope with the careers of Innocent III and Bill Gates. [1]

Yet, I shall argue, changes such as the transformation of mediaeval Europe under popes like Innocent III and the current profound restructuring of industrial societies at the hands of software entrepreneurs like Bill Gates are explained, and indeed predicted, in the writings of the Canadian economic historian Harold A. Innis.  During a remarkable half-decade of research between 1945 and 1950, Innis found what he believed to be a recurrent pattern in western social history.  He expressed his ideas in such enigmatic, even oracular, fashion, however, that readers of his own day and since have had difficulty understanding them.  As Marshall McLuhan later wrote, Innis’s writing in this period is compressed into a ‘mosaic structure of seemingly unrelated sentences and aphorisms’ (McLuhan 1964, vii).  Three samples (with italics added) will suffice to give the flavour of his style.  ‘[The introduction of monopolistic elements in culture] is accompanied by... collapse in the face of technological change which has taken place in marginal regions’ (1951, 4).  ‘The tenacity of the Byzantine empire assumed the achievement of a balance which recognized the role of space and time’ (1950, 167).  ‘The relative emphasis on time or space will imply a bias of significance to the culture in which [a medium of communication] is imbedded’ (1951, 33).  ‘Marginal regions,’ ‘balance,’ and ‘bias’: what does Innis mean?  These quotations are taken from three lectures that Innis gave in 1947 at Laval University, in 1948 at Oxford University, and in 1949 at the

1. Lucas (1988) and Romer (1986, 1990) have been among the most innovative in addressing the problem of endogenizing technical change within the neoclassical model, allowing progress to depend on the stock of accumulated knowledge.

755

University of Michigan.  In each case, Innis’s method was to look for patterns in the historical flows of names, events, and dates.  Although there were no equations or logical proofs in his writing, he did, I shall argue, have a theory.  To understand Innis’s ideas, however, we must decode the rather special vocabulary with which he expressed his thoughts.  In doing so, we shall compare Innis’s theory with the generalized Darwinian model of evolution as defined by the American psychologist, Donald T. Campbell (1965, 27). [2]   In addition, we shall explore the links between Innis’s ideas and the essential elements of recent evolutionary modelling in economics, as set out by the German economist Ulrich Witt (1991). [3]  Finally, we shall examine Innis’s theory of communications as the key to a new approach for modelling technological change.

II. The Margin

In the late 1940s and 1950s, Harold A. Innis was perhaps Canada’s leading scholar.  He was the author of three well-received studies on Canadian economic history, A History of the Canadian Pacific Railway (1923), The Fur Trade in Canada (1930), and The Cod Fisheries (1940).  In each book, Innis examined how changes in the techniques for producing primary products - staples - affected patterns of social organization over long periods of time in regions on the periphery of European civilization.  He was head of the Department of Political Economy at the University of Toronto and was first dean of the graduate school of that university.  As a sign of international recognition, he was elected president of the American Economic Association.

In 1945, immediately following the Second World War, Innis was invited to visit the Soviet Union.  He was fifty years old at that time, and the journey marked a turning point in his thinking.  It was a shock for him to witness the accomplishment of a structure of social organization so completely different from the market system that he had studied in detail.  In a letter, he wrote: ‘I felt the necessity for a much broader approach in economic history and the very great danger of a very narrow approach such as we seem to get nothing else but’ (Creighton 1957/1978, 122).

The opportunity to deliver a more general message to a wider audience came the following year.  In May 1946 Innis was elected president of the Royal Society of Canada, an interdisciplinary group of leading Canadian researchers.  The position required that he deliver a presidential address the following year.  It was about this time that Innis began to put together what we might today call a set of hypertext files.  They were cross-referenced index cards on which Innis noted his ideas as they came to him. [4]

2. Durham (1991) reviews the development of Darwinian theories in anthropology.

3. Robson (1995) applies an evolutionary model to explain attitudes towards risk in economics; Carmichael and MacLeod (1994) suggest that social customs may be modelled with biological models; Paquet (1995) surveys theories of institutional evolution.

4. These notes were subsequently sorted and published by Christian (1980).

756

Innis presented a first version of his broader approach to economic history in May 1947 in Quebec City in a lecture entitled, ‘Minerva’s owl’ (1951, chap. 1).  The theme of his study was the diffusion of new technologies - how the use of an innovation spreads over time.  This subject had been discussed in brilliant fashion in the recent writings of Joseph Schumpeter, who explained cyclical changes in output by the impact of innovations. In Capitalism, Socialism and Democracy (1942), the Austrian economist had argued that the principal source of contemporary economic innovation was the large-scale corporation, which alone had the capacity to coordinate research and the resources to apply new technologies.

Innis’s ideas were sharply different from those of Schumpeter.  Instead of dealing with new technologies in production and transportation, Innis focused on new communications media.  Rather than limit himself to markets for goods and services, he enlarged his scope to cover the overall pattern of interaction within the society.  Instead of concentrating on the monopolistic structures at the centre, he directed his attention to the competitive fringe, the contesting forces on the geographic periphery.

The question that Innis asked his Quebec audience was the following.  If the dominant group within an organization derives its power from a monopoly of the existing communications technology, how can an alternative form of communication with different characteristics possibly spread?  Taking issue with Schumpeter, he argued that it is not in the interest of the group that holds power to encourage innovation.  Indeed, he went on, those who hold a monopoly of knowledge will actively resist the introduction of new techniques.  He quoted Albert Guérard: ‘To the founder of a school, everything may be forgiven except his school’ (1951, 4).  As a result, a new technology can spread only in marginal regions where the dominant elite’s power is weak.  ‘In the regions to which Minerva’s owl takes flight the success of organized force may permit a new enthusiasm and an intense flowering of culture incidental to the migration of scholars engaged in Herculean efforts in a declining civilization to a new area with possibilities of protection’ (1951, 5).

As in most of his last works, Innis’s style was that of theme and historical variations, with very little in the way of recapitulation.  After a brief introduction written in this dense, epigrammatic style, Innis proceeded to a long series of historical case studies.  One of the most interesting of these examples concerns the spread of parchment, along with a new script and a standardized written vehicular language, in early mediaeval Europe.  Parchment, a writing material made from the skins of animals, was discovered about 200 BC.  Until the eighth century, however, the preferred medium for administrative purposes was papyrus, a kind of paper made from reeds.

Within the Christian Church, a monastic movement had begun by the fourth century.  Under the later Roman emperors and their Byzantine and Germanic successors, the monasteries remained under centralized control.  Only on the fringes of the former Roman Empire, in Ireland and subsequently in England, did the monasteries acquire a certain degree of autonomy.  Here the abbots took the initiative of having existing works from throughout Christendom transcribed onto parchment.

757

With the Arab conquest of much of the Mediterranean basin in the seventh and eighth centuries, supplies of papyrus to western Europe were cut to a bare trickle.  Since papyrus decomposes within about three generations, virtually the only knowledge retained was that transferred to much more durable parchment by monks.  The parchment codex manuscripts in the libraries of English monasteries therefore became a principal repository of European learning.  From England St Boniface (675-754) was sent out to convert the pagan German tribes, introducing Latin on parchment into central Europe.  Another Anglo-Saxon churchperson, Alcuin (735-804), was the principal adviser to Charlemagne and the author of reforms that standardized Latin culture across western Europe (ibid., 14-17).  Innis described how the ‘clear, precise, and simple’ Carolingian minuscule, developed under Alcuin at the court of Charlemagne, gradually spread throughout western Europe.

In this example, Innis was explaining how a new communications medium, standardized Latin as a vehicular language in a new script on a new material, gradually spread from the British Isles throughout northern Europe.  In essence, the new idea was carried from monastery to monastery from the periphery to the centre, successfully replicating itself in each new host institution.

Is there a more general process involved?  In an essay on theories of cultural evolution published in 1965, Donald T. Campbell set out the criteria that a Darwinian model of evolution must satisfy.  One of them is ‘a mechanism for the preservation, duplication, or propagation of... variants’ (1965, 27).  Innis’s concept of the margin, by which a new communications technique establishes a foothold in an isolated region, successfully reproduces itself, and eventually challenges the dominant medium, would appear to satisfy this criterion.  Indeed, the suggested mechanism is quite similar to that proposed by biologists to explain the emergence of new species (see, e.g., Dawkins 1986).

Why should Innis’s approach be of interest to economists?  The economic analogue to the biological concept of reproduction is the process of technological diffusion.  Witt (1991, 91-6) points out that the basic principle in evolutionary models of diffusion is what is called the frequency-dependency effect.  Each individual in a population must decide whether or not to adopt a new technique.  The probability that she adopts, however, will be a function of the number of members of the population who have already adopted it.  In the mediaeval case, monks travelling slowly from one monastery to another were the principal means of diffusion.  The greater the number of monasteries that had already adopted the technology, the more likely a non-adopter was to receive a visit from converts to the innovation.  In any given sub-population of monasteries, however, some threshold may have been required before the new technique was adopted.

In short, in this essay Innis took the first step towards the construction of an evolutionary model of social change.  His theory was still incomplete, but already Innis had left his audience of the Royal Society of Canada far behind.  His biographer, the historian Donald Creighton, wrote that his lecture, ‘Minerva’s owl,’ was ‘much too long’ and that ‘many in his audience were puzzled or bewildered’ (1957, 127).

758

III. Balance

In June 1946, a month after his election as president of the Royal Society of Canada, Innis had received an invitation from the administrators of the Beit Fund of Oxford University.  They asked him to deliver six lectures ‘on any subject in the economic history of the British Empire,’ undoubtedly expecting him to write on the economic relations of colonial regions like Canada to the imperial centre in Britain (Creighton 1957, 126).  Over the next two years, Innis did indeed prepare six lectures for an international audience.  The context of these lectures was by no means confined to a single subject, however, nor did he limit his attention to economic phenomena; most strikingly, his talks had little to do with the British Empire.

Innis began the Beit Lectures on 12 May 1948 at All Souls College.  He warned his listeners at the outset that a slight digression might be necessary.  ‘I shall attempt to outline the significance of communication in a small number of empires as a means of understanding its role in a general sense and as a background to an appreciation of its significance to the British Empire.’  Six lectures, later, this slight digression to look at a small number of empires turned out to have been a new theory of historical change.

We have seen that the first of Innis’s concepts, the margin, proposed a process of technological diffusion.  The question that he now addressed was whether or not this process attained an equilibrium and if so, whether two or more technologies would coexist in such a state. [5]  For each communications medium, Innis asserted, there will be a corresponding form of social organization.  Some media will favour decentralization, while others will favour centralization.  Balance is reached when the centrifugal forces of the former are exactly offset by the centripetal forces of the latter.  ‘Large-scale political organizations such as empires... have tended to flourish under conditions in which civilization reflects the influence of more than one medium’ (1950, 7).  It is in such a situation, Innis felt, that social welfare will be highest.  We have seen that he noted approvingly the balance between what he referred to as time and space in the Byzantine Empire.

A more likely outcome was for one medium to dominate the others.  Such a monopoly situation would then lead to rigidity and decay.  ‘We can perhaps assume that the use of a medium of communication over a long period will eventually create a civilization in which life and flexibility will become exceedingly difficult to maintain’ (ibid., 34).

It was such failure to achieve balance, Innis argued, that led to the political instability characteristic of the west in pre-modern times.  Innis devoted an entire lecture in his Oxford series to mediaeval Europe.  Under the papyrus and stylus medium of the Roman Empire at its height, Europe was administered from a central point.  But under the parchment and pen medium that succeeded it, power was decentralized.  The hierarchic structures of empire were replaced by networks

5. In terms of evolutionary theory, does the process of replication lead to an evolutionarily stable equilibrium in which mutant replicators are eliminated?  If so, is that equilibrium polymorphic, that is, characterized by more than one replicator?  See Binmore (1992, 422-9).

759

of monasteries and their secular overlords.  ‘In contrast with papyrus, which was produced in a restricted area under centralized control to meet the demands of a centralized bureaucratic administration and which was largely limited by its fragile character to water navigation, parchment was the product of a widely scattered agricultural economy suited to the demands of a decentralized administration and to land transportation’ (ibid., 140).

In western Europe, a brief period of balance between the state with its tendency towards centralization and religion with its generally decentralized structure was achieved during the Carolingian renaissance from 768-814.  After Charlemagne’s reign, however, the empire of the Franks broke up into competing states, the basis of the modern nations of Europe.  By the thirteenth century the papacy under leaders like Innocent III had managed to impose its authority over the secular rulers of these successor states, including Frederick II, the Holy Roman Emperor.  ‘A civilization dominated by parchment as a medium developed its monopoly of knowledge through monasticism.  The power of the Church was reflected in its success in the struggle with Frederick II’ (ibid., 165).

What is the process by which one medium and its associated form of social organization are replaced by another medium and an alternative social structure?  For Innis, one mechanism was competition for territory between states.  For example, even though it was much smaller in territory than the Roman Empire, the kingdom of Charlemagne was too large to survive external predation.  ‘Attacks from the Danes and the Magyars accentuated local organizations of force and separatist tendencies’ (ibid., 149).  A second mechanism was competition between centre and periphery for popular support within a given society. ‘ The monopoly of knowledge which had been built up invited competition from a new medium of communication which appeared on the fringes of western European culture and was available to meet the demands of lower strata of society’ (ibid.).

In short, a system that is more successful than another in feeding and protecting the majority of its adherents will tend to be selected.  This process by which competition between systems results in retention of one way of life and elimination of the other in Innis’s theory therefore satisfies the second of the requirements set out by Campbell for a Darwinian model of evolution; namely, ‘consistent selection criteria’ (1965, 27).  A struggle for survival among states and among groups within states weeds out less successful forms of organization.

Innis’s idea of balance should be distinguished from the concept of equilibrium in neoclassical economic theory.  As Witt (1991, 96-8) has remarked, neoclassical general equilibrium theory takes a strong position on the issue of social coordination, proposing that an economic system converges to a state of perfect coordination - equilibrium.  All firms in an industry, for example, will use the identical production technology.  For an evolutionary approach, it is necessary to introduce the de-coordinating force of innovative activity.  To survive, agents must keep within the moving bounds set by coordinating forces of markets and the de-coordinating forces of innovation.

By these standards, Innis’s theory is clearly evolutionary.  At any moment in a given society, different communications technologies are likely to coexist.  Indeed,

760

his ideal is a state of balance in which two or more media of communication are used.  Such a state of balance is likely to be temporary, Innis realizes, since one technology will tend eventually to dominate the other.  Even then, however, there is no enduring equilibrium, since, as we have seen, the monopoly position of the dominant technology induces de-coordinating innovative activity.

Once again, Innis had raced far ahead of his audience.  The reception of the six Oxford lectures also was disappointing.  ‘Innis’s most comprehensive and original thesis was presented prematurely, too briefly and without the [necessary] vast mass of supporting evidence and illustrative material’ (Creighton 1957, 135).  Yet he had accomplished the essential.  In his two years of research, Innis had acquired a mastery of the historical details of social change.  He could now add the third and crucial element to his theory.

IV. Bias

During the winter of 1948-9, while Innis was finishing the revisions to his Oxford lectures, to be published as Empire and Communications (1950), he received an invitation to participate in a Royal Commission on Transportation headed by W.F. Turgeon.  The gruelling task would consume much of his energy over the following two years.  But before becoming involved in the preparations for the Commission’s hearings, Innis took the time to prepare an essay, ‘The bias of communications,’ which he delivered at the University of Michigan in April 1949.  The result was a synthesis of the two previous studies and his most complete explanation of the process of historical change.

Perhaps the most difficult of all problems in economic theory is to explain how new ideas arise.  To deal with this problem, Innis used the concept of bias. [6]  Here Innis was dealing with the direction of technical change.  He was not, as one might expect, referring to whether capital or labour is saved, but rather to whether an innovation is designed to allow knowledge to be transmitted more efficiently over space or to be preserved more effectively over time.  ‘According to its characteristics [a medium of communication] may be better suited to the dissemination of knowledge over time than over space, particularly if the medium is heavy and durable and not suited to transportation, or to the dissemination of knowledge over space than over time, particularly if the medium is light and easily transported’ (1951, 33).  Innis suggested that both the direction and the rate of change are endogenously determined.  ‘Monopolies or oligopolies of knowledge have been built up in relation to the demands of force chiefly on the defensive, but improved technology has strengthened the position of force on the offensive and compelled realignments favouring the vernacular’ (ibid., 32).  In other words, the elite group that dominates the old technology will receive a monopoly rent from its position.  However, the high price for information will encourage the development

6. Innis also used the notion of bias in the introduction to Empire and Communications (1950, 7).  The idea that the bias of the dominant medium leads to innovation to correct the bias, however, is expressed most clearly and consistently in the 1949 paper (1951, 34, 38, 40, 48, 49, 50, 60).

761

of a new technology with different characteristics, intended to lower the cost of knowledge to the rest of the society.

An example from the mediaeval period illustrates this concept of bias and its significance.  The communications medium of the Roman Empire, light but perishable papyrus, encouraged an excessive concern for territory; that is, space.  In the eighth century the Arab conquest of the south shore of the Mediterranean Sea drove up the price of papyrus to western Europe.  The Carolingian monks were therefore forced to turn to an alternative medium, parchment, which was heavier but much more durable.  Combining parchment with a new standardized script, the Carolingian minuscule, they developed a powerful new communications medium.  As a result, over the following centuries the previous bias towards extension over space was replaced by a bias towards duration over time.  The military-administrative bureaucracies of the Roman Empire and its successor states were replaced by the religious bureaucracies of the high middle ages (ibid., 48-50).

In proposing a process of directed endogenous technological change, Innis was taking a position sharply at variance with neoclassical economics, which has considered technology to be exogenous.  However, he was offering the final component of an entirely different approach to social organization.  The third of Campbell’s criteria for a Darwinian model of evolution is the ‘occurrence of variations’ (1965, 27).  Innis’s concept of bias, by which monopoly pricing generates new technology with characteristics different from the old, clearly fits this requirement.

Innis went beyond the Darwinian model.  The mutations in technology he described are not the results of a random process.  Rather, they are a reaction to price data.  Innovations seem to be generated by what Witt (1991, 89) calls a ‘satisficing’ mechanism in which action arises from dissatisfaction with the current outcome.  If so, Innis was going beyond the evolutionary theory of his contemporary, Schumpeter, who had little to say about how new ideas arise (Freeman 1990, 22).  He was proposing the key element of a theory of social change, namely, a model of cultural evolution.

There is no indication in Creighton’s biography of how his American audience reacted to this presentation.  Together with Innis’s 1947 Royal Society address and several other essays, the Michigan lecture was published in 1951 as a book, The Bias of Communication.

V. A Bumper-Car Theory of History

A popular ride in amusement parks of the early postwar decades was the electric bumper car.  By pressing on a pedal, the rider caused the car to move at a fixed velocity until it hit an obstacle - either the bumper of another car or the barrier around the concession.  The vehicle then changed direction until it hit some other obstacle.  Innis’s theory of social change somewhat resembles this amusement.  The position of the vehicle represents the nature of communications technology.  As for the car, it represents social institutions, and the rider, the population of the society, each carried along by the momentum of technological change.  The bumpers are

762

the limits to the price of storing information over time as opposed to the price of transmitting it over distance.  At these limits there is a redirection of society’s resources to search for an alternative communications medium.

Innis’s theory of history, as told in condensed form in his essay, ‘The bias of communication’ (1951, 33-60), may be illustrated with the help of figure 1.

Here, the ratio of the cost of information storage to the cost of information transmission is measured on the vertical axis, while the year is measured on the horizontal axis.  In the prehistoric period, as information accumulated with a fixed memory capacity for each individual, the marginal cost of storing a unit of information reached the upper limit.  In Mesopotamia, a way was found of storing information by means of pictograms on clay tablets.  Subsequently, phonetic writing developed, using the same materials.  The first civilizations, ruled by a priestly caste, were concerned with time.

The information monopoly of the ruling elites encouraged the search for alternative media.  With the development of papyrus, it became possible to send information efficiently over long distances.  The cost of transmission was further reduced when the west-Semitic peoples developed a consonantal alphabet capable of representing human speech with two dozen characters.  Using certain of these symbols for vowels, the Greeks and Romans developed complete alphabets and came to dominate all of the lands in the Mediterranean basin.  Secular rulers with codes of written law now emphasized the control of space.

Since papyrus crumbled after three generations, it was not suitable for conserving information over long periods.  In the early Christian era, the development of durable parchment, which could be bound into easily consultable books provided

763

an alternative means of communication.  When Arab conquests caused the price of papyrus to rise in the seventh and eighth centuries, western Europe switched to parchment and developed a standardized efficient Latin script.  European society of the middle ages was dominated by a religious elite, literate in this vehicular medium and concerned with the preservation of knowledge over time.

From the fourteenth century on, the high cost of sending information over space encouraged the European production of paper, an invention borrowed from the Chinese.  Once again, a lighter, perishable medium was being substituted for a heavier, durable one.  In the fifteenth century the high price of hand-copied manuscripts provided an incentive to experiment with mechanical methods of reproducing information.  The resulting fall in transmission costs led to the development of written forms of vernacular languages and intense competition for territory.  In the nineteenth century the invention of newsprint and steam and electric presses further marked the transition to media adapted to the control of space.  In the twentieth century radio continued this trend towards an obsession with the present.

We are now nearing the moment when Innis’s career itself came to an end.  What might he have predicted for the remainder of the twentieth century?  Applying the model of figure 1, he might well have suspected that after some seven centuries of preoccupation with space, the bumper car should hit the upper limit of the ratio of storage of transmission costs.  Accordingly, innovation to reduce storage costs would be induced.  The consequence would be a new concern for time and difficulties with any existing institutions committed to the control of space.

Applying this reasoning, Innis might have predicted fiscal problems for existing territorial states.  Difficulty in maintaining cohesion might be particularly severe for states like the Soviet Union, which had hitherto neglected the time dimension or for states like Yugoslavia and Canada in which relatively homogeneous regions had different concepts of time.  In addition, Innis might have foreseen additional strength for movements within individual societies devoted to furthering moral values.  This is exactly the pattern that has been observed in the industrialized world since the introduction of a powerful new device for storing information, the integrated circuit, a decade after Innis’s death. [7]  Bill Gates is one of those who have recognized its revolutionary implications.  It is the same pattern found in the history of mediaeval Europe after the death of Charlemagne.  Indeed to foresee where we might be headed, we might do well to study the career of Innocent III, who was not even an ordained priest when elected pope at the age of thirty-seven yet went on to reshape European society.

VI. The Missing Dimension: Number

It might be argued that it is absurd to attempt to fit all of western social history within a simple, two-dimensional view of communications technology as agent of social change.  Indeed, the strongest arguments against the time-space bumper-car

7. For more on this new technology and its effects on patterns of social organization, see Dudley (1991, chap. 8).

764

model of the preceding section were provided by Innis himself.  In the introduction to one of his last essays, ‘The problem of space,’ he briefly mentioned an additional concept that he placed on the same level as time and space, namely, number.  ‘Gauss held that whereas number was a product of the mind, space had a reality outside the mind whose laws cannot be described a priori.  In the history of thought, especially of mathematics, Cassirer remarked, “at times, the concept of space, at other times, the concept of numbers, took the lead”’(1951, 92; italics added).

Could number be considered a third dimension to communications technology?  If so, how might this concept fit into the structure that Innis designed to model space and time?  At issue, Innis realized, is the complexity of the communication system.  ‘A complex system of writing becomes the possession of a special class and tends to support aristocracies.  A simple flexible system of writing admits of adaptation to the vernacular but slowness of adaptation facilitates monopolies of knowledge and hierarchies’ (ibid., 4).  Thus when a complex system of communication is replaced by a simpler one, there is deeper penetration into the society.  What was formerly reserved for an elite becomes accessible to a much wider segment of the population.

As a result, there is a change in the nature of social interaction.  Innis described two periods in history when the number of users of a communications medium increased dramatically.  The first change came with the invention of the alphabet in the eastern Mediterranean in the second millennium BC.  With the development of a flexible alphabet, the nature of society changed.  ‘A flexible alphabet favoured the growth of trade, development of trading cities of the Phoenicians and the emergence of smaller nations dependent on distinct languages’ (ibid., 39).  Innis noted the divisive effects of the new technology within the Roman Empire, which towards its end separated into Greek and Latin components (ibid., 15).

The second period in which number became crucial occurred subsequent to the development of printing in standardized vernacular languages in early-modem Europe.  ‘By the end of the sixteenth century the flexibility of the alphabet and printing had contributed to the growth of diverse vernacular literatures and had provided a basis for divisive nationalism in Europe’ (ibid., 55).

In each case, Innis recognized the nature of the change but was unable to develop the analytical structure required to treat it theoretically.  As a result, number became confused with space in his analysis.  It is only after Innis’s death, with the development of the theory of the public good by Samuelson (1954) and its application to interest-group activities by Olson (1965), that the implications of number became apparent.  The larger the number of contributors to the financing of a public good, the lower the price to each.  To the extent that the institutions required to support a national written language form a public good, the taxpayers of each nation will therefore have an interest in increasing the numbers of its citizens.

Indeed, there is an additional significance to number.  Increasing number not only lowers the costs but also raises the benefits of information exchange in the vernacular.  When a new member joins such a system of two-way communications, she confers a benefit - a network externality - to all existing members of the system, since they can now communicate with an additional user (Katz and Shapiro

765

1985).  Thus number is a crucial characteristic of certain media of communication.  Here, then, is a task for those who would write the final movements of Innis’s unfinished symphony: the integration of number into a model of endogenous change in information technology.

Why should change in information technology command special attention at the expense of more traditional areas of interest to economists, such as agricultural techniques, the development of inanimate power sources, or new means of transport?  In a word, communications media are arguably the most fundamental of ‘enabling technologies’ - techniques that allow us to use other means of production. [8]  Rather than simply providing us with more goods and services from given inputs, innovations in information processing tend to lead us to interact differently.  Recent research on primate evolution suggests that greater memory, an increased ability to communicate over distance, and a fall in the cost of reproducing information through speech each have had a profound impact on patterns of social organization (see, e.g., Ghiglieri 1989; Leakey and Lewin 1992; Johanson, Johanson, and Edgar 1994).  What is new in the historical period analysed by Innis is that these changes are external to the individual - the result of cultural rather than biological evolution.

VII. Conclusion

In November 1952 Harold Innis died of cancer at the age of fifty-eight.  His fellow economists, W.T. Easterbrook (1960) and Mel Watkins (1963), along with Marshall McLuhan (1962), subsequently paid tribute to him in their writings in the early 1960s.  Since then, Innis’s memory has been kept alive by researchers in other disciplines, who recognize his contributions to the theory of communications, political science, and geography. [9]  Among economists and economic historians, however, he is rarely cited. [10]  Is there something that economists have missed?

As they have grown older, many of the most active of later twentieth-century economists have quietly dropped the assumption of rational, optimizing agents learned in their youth.  Friedrich Hayek (1973-9), Kenneth Boulding (1973), Jack Hirshleifer (1980), Richard Nelson (1987), Douglass North (1990), Nathan Rosenberg (1994), and Richard Lipsey (1995) are among the more prominent of those who have become dissatisfied with the limits of neoclassical economic theory.  Seeking both a more satisfactory way of modelling technological change and a means of broadening the scope of their analysis to cover cultural phenomena, they have been attracted by the concept of evolution.  All of these researchers appear to have been preceded by Harold A. Innis, although none has cited Innis’s work on communication in his own writings.  The basic concepts of Innis’s theory of

8. For an analysis of this concept, see Lipsey and Bekar (1995).

9. For a review of research on Innis’s ideas on media, see Di Norcia (1990). Christian (1977) portrays Innis as political scientist, Parker (1988) as geographer.  More recently, Innis’s writings on media have been cited in the discipline of communications by Geiger and Newhagen (1993) and Deetz (1994).

10. Neill (1972) and Parker (1988) are exceptions among economists.

766

social change - bias, balance and the margin - are simply new terms for the Darwinian concepts of mutation, selection, and reproduction.  In addition, Innis pointed to the importance of communications media as the most fundamental of enabling technologies.

Yet there is a fundamental objection to this interpretation of Innis’s later work as evolutionary economics.  At no point did Innis himself draw a parallel between Darwin’s concept of evolution as ‘descent with modification’ and his own theory of social change.  Innis’s posthumously published system of filing cards indicates that he was aware of Darwin’s impact on nineteenth-century social scientists like Herbert Spencer, who coined the term ‘survival of the fittest’ (Christian 1980, 7, 39).  A possible explanation for his own reticence to employ Darwinist terminology is that in the late 1940s Darwinism in the social sciences had received an ugly reputation as a result of racist theories of natural selection advanced to justify Nazi brutality.

Another interpretation of Innis’s silence on this question is also possible.  Innis may have viewed both Darwin and himself as continuing a much older tradition dating back to the humanism of the sixteenth century.  In his essay, ‘A plea for time,’ the third chapter of his 1951 book, Innis wrote: ‘The linear concept of time was made effective as a result of humanistic studies in the Renaissance... It was not until the Enlightenment that the historical world was conquered and until Herder and romanticism that the primacy of history over philosophy and science was established... In the hands of Darwin the historical approach penetrated biology and provided a new dimension of thought for science (62-3).

Innis, like Darwin and Marx, was proposing a historical, path-dependent theory of change.  In this sense he was very much a follower of the German idealist philosopher G.W.F. Hegel, for whom history was a dialectic process by which conflict between opposing forces is resolved.  The title of Innis’s 1947 Quebec address is drawn from the preface to Hegel’s The Philosophy of Right (1952): ‘When philosophy paints its grey on grey, then has a shape of life grown old.  By philosophy’s grey on grey it cannot be rejuvenated but only understood.  The owl of Minerva spreads its wings only with the falling of the dusk’ (7).  Hegel was saying that it is only in the final phase of a civilization, when decline is irreversible, that its culture begins to flower.  But Hegel’s statement, written late in his own life, could also apply to a individual thinker like himself - or Harold Innis.  Innis’s great synthesis came late in his life when his hair was streaked with grey.  He died convinced he had found the means of understanding the nature of his own civilization.  But he himself was by that time powerless to influence the thinking of his contemporaries.

Today, with the technology to cram much of world’s learning onto an inexpensive disk available to children, centralized information storage is becoming obsolete.  Our increasing focus on the environment, social justice, and spiritual values may be explained, at least in part, by our cheaper access as individuals to the information we need to behave consistently over time.  After seven centuries of what Harold Innis considered an increasing obsession with space, the west appears to be moving

767

in the direction of a more balanced society.  The moment perhaps has come for Minerva’s owl to alight on the shoulders of a new generation.

References

Binmore, K. (1992) Fun and Games: A Text on Games Theory (Lexington, MA: Heath)

Boulding, K.E. (1973) The Economy of Love and Fear (Belmont, CA: Wadsworth)

Campbell, D.T. (1965) ‘Variation and selective retention in sociocultural evolution.’ In Social Change in Developing Areas: A Reinterpretation of Evolutionary Theory, ed. H.R. Barrington, G.I. Blanksten, and R.W. Mack (Cambridge, MA.: Schenkman)

Carmichael, H.L. and W.B. MacLeod (1994) ‘Gift giving and the evolution of cooperation.’ Queen’s University and Université de Montréal Working Paper

Christian, W. (1977) ‘Harold Innis as political theorist.’ Canadian Journal of Political Science 10, 21—42

_________ . (1980) The Idea File of Harold Adams Innis (Toronto: University of Toronto Press)

Creighton, D. (1951/1978) Harold Adams Innis: Portrait of a Scholar (Toronto, University of Toronto Press)

Dawkins, R. (1986) The Blind Watchmaker (New York: Norton)

Deetz, S.A. (1994) ‘The micro-politics of identity formation in the workplace: the case of a knowledge intensive firm.’ Human Studies 17, 23-44

Di Norcia, V. (1990) ‘Communications, time and power: an Innisian view.’ Canadian Journal of Political Science 23, 335-57

Dudley, L. (1991) The Word and the Sword: How Techniques of Information and Violence Have Shaped Our World (Cambridge, MA: Basil Blackwell)

Durham, W.H. (1991) Coevolution: Genes, Culture and Human Diversity (Stanford: Stanford University Press)

Easterbrook, W.T. (1960) ‘Problems in the relationship of communication and economic history.’ Journal of Economic History 20, 134-5

Freeman, C. (1990) ‘Schumpeter’s Business Cycles revisited.’  In Evolving Technology and Market Structure: Essays in Schumpeterian Economics, ed. A. Heertje and M. Perlman (Ann Arbor: University of Michigan Press)

Geiger, S., and J. Newhagen (1993) ‘Revealing the black box: information processing and media effects.’ Journal of Communication 43, 42-50

Ghiglieri, M.P. (1989) ‘Hominid sociobiology and hominid social evolution.’ In Understanding Chimpanzees, ed P.G. Heltne and L.A. Marquardt (Cambridge, MA: Harvard University Press)

Campbell, D.T. (1965) ‘Variation and selective retention in sociocultural evolution.’  In Social Change in Developing Areas: A Reinterpretation of Evolutionary Theory, ed. H.R. Barrington, G.I. Blanksten, and R.W. Mack (Cambridge, MA: Schenkinan)

Hayek, F.A. (1973-9) Law, Legislation and Liberty. 3 vols. (London: Routledge & Kegan Paul)

Hegel, G.W.F. (1952) The Philosophy of Right (Chicago: Encyclopedia Britannica)

Hirschleifer, J. (1980) ‘Privacy: Its origin, function and future.’ Journal of Legal Studies 9, 649-64

Innis, H.A. (1923) A History of the Canadian Pacific Railway (Toronto: McClelland and Stewart)

_______ . (1930) The Fur Trade in Canada (Toronto: University of Toronto Press)

_______ . (1940) The Cod Fisheries (Toronto: University of Toronto Press)

_______ . (1950) Empire and Communications (Oxford: Clarendon)

_______ . (195 1/1964) The Bias of Communication (Toronto: University of Toronto Press)

768

Johanson, D., L. Johanson, and B. Edgar (1994) Ancestors: In Search of Human Origins (New York: Villard)

Katz, M.L., and C. Shapiro (1985) ‘Network externalities, competition, and compatiblity.’ American Economic Review 75, 424-40

Leakey, R., and R. Lewin (1992) Origins Reconsidered: In Search of What Makes Us Human (New York: Doubleday)

Lipsey, R.G., and C. Bekar (1995) ‘Technical change and economic growth: continuous random shocks vs. occasional paradigm shifts.’ In Technology, Information and Public Policy, ed. T.J. Courchene (Kingston, Os: John Deutsch Institute), forthcoming

Lucas, R.E. Jr (1988) ‘On the mechanisms of economic development.’ Journal of Monetary Economics 22, 3-42

McLuhan, M. (1962) The Gutenberg Galaxy: The Making of Typographic Man (Toronto: University of Toronto Press)

_________ . (1964) ‘Introduction.’ In The Bias of Communication (Toronto: University of Toronto Press)

Neill, R.F. (1972) A New Theory of Value: The Canadian Economics of H.A. Innis (Toronto: University of Toronto Press)

Nelson, R.R. (1987) Understanding Technical Change as an Evolutionary Process (Amsterdam: North-Holland)

North, D.C. (1990) Institutions, Institutional Change and Economic Performance (Cambridge: Cambridge University Press)

Olson, M. (1965) The Logic of Collective Action (Cambridge, MA: Harvard University Press)

Paquet, G. (1995) ‘Institutional evolution in an information age.’ In Technology, Information and Public Policy, ed. T.J. Courchene (Kingston, ON: John Deutsch Institute), forthcoming

Parker, I. (1988) ‘Harold Innis as a Canadian geographer.’ Canadian Geographer 32, 63-9

Robson, A.J. (1995) ‘The evolution of strategic behaviour.’ This Journal 28, 17-41

Romer, P.M. (1986) ‘Increasing returns and long-run growth.’ Journal of Political Economy 94, 1002-37

_________ . (1990) ‘Endogenous technological change.’ Journal of Political Economy 98, S71— S102

Rosenberg, N. (1994) Exploring the Black Box (Cambridge: Cambridge University Press)

Samuelson, P.A. (1954) ‘The pure theory of public expenditure.’ Review of Economics and Statistics 36, 387-9

Schumpeter, J.A. (1942) Capitalism, Socialism and Democracy (New York: Harper)

Watkins, M.H. (1963) ‘A staple theory of economic growth.’ Canadian Journal of Economics and Political Science 29, 141-58

Witt, U. (1991) ‘Reflections on the present state of evolutionary economic theory.’ In Rethinking Economics, ed. G.M. Hodgson and E. Screpanti (Aldershot, England: Edward Elgar)

769