SOURCE: "Rationality and the Informational Environment: A Reassessment of the Work of Harold Adams Innis," in Journal of Canadian Studies/Revue d'études canadiennes, Vol. 22, No. 4, Winter, 1987-88, pp. 78-92.

A NEW PERSPECTIVE

Harold Innis's message was not well communicated, because, in part, those who received it did not occupy a perceptual vantage point from which it could be understood. The consequence has been a multiplication of interpretations of Innis, many cited in R.F. Neill's intellectual biography,¹ others appearing since 1972 in the continuing flow of commentaries. None has captured Innis's fundamental insight into the nature and consequences of the informational environment of decision-making.

Over the past fifteen years, with no reference to Innis—perhaps because they have misunderstood him, perhaps because they have never heard of him—a number of economists have taken a new approach in which Innis's work makes consistent sense. In the consequent perspective, his general research program may appear to be clearly separated from the sequence of subjects to which it applies and with which it deals. Indeed, in this light, the underlying unity of his economic analysis, his Canadian history, and his essays on communication becomes increasingly evident. The methodological issues at the root of his studies, his general scientific contribution, and his innovative interpretation of the Canadian economy all come to the fore.

The new approach in economics was undertaken by the proponents of New Classical Economics in their attempt to single out and negate the principal tenets of Keynesian theory. The broad Keynesian judgement—that there is something wrong with the market system that has to be corrected by government—had to be effectively challenged by a neoconservative assertion that just the opposite is the case, that is, that there is something wrong with government intervention that has to be corrected by increased reliance on the market system. In generating this challenge, the new conservatives replaced the theory of irrational expectations that was the foundation of the Keynesian interpretation with a theory of rational expectations. This, in turn, triggered a renewed awareness that a theory of the informational environment—that is, a theory of the decision-making process—is central to all economic analysis.

A few years before Innis's death in 1952, empirical development of the theory of decision-making accelerated under the influence of Herbert Simon and his associates.² Like Innis's theorizing, Simon's work has not been accepted within the mainstream of economics. Nonetheless, it has had a regenerating influence on the discipline because of its more analytic approach to consumer and producer choice. It is a basic and inductive, rather than an applied, approach. It focuses more narrowly on the process of deciding and its consequences than on the expectant outcomes of the decision-makers. A listing of Simon's conclusions helps to put Innis's contribution into an appropriate perspective. According to Simon there are specifiable limits or "bounds" to the use of reason in decision-making. These are: (1) the limits of computational ability, set by the limits of short-term and longterm memory; (2) the limits of perception, set by the path of information search; (3) the limits set by the availability and certitude of information; (4) the limits set by the presence of multiple goals; and (5) the limits set by preoriented attention. Innis was particularly concerned with the historical effects of the bounding of rationality by limited perception and attention.

An important conclusion drawn by Simon is that the rational, maximizing economic man as assumed in neoclassical economics has no empirical validity. Agents are, as he puts it, "satisficing" rather than maximizing. Innis, who was innocent of empirical psychology, never used the term "satisficing," but he agreed with the substance of this proposition,³ and he sought with a different kind of empiricism to specify its consequences in economic behaviour. He prefaced his essays on the bias of communication with the comment, "They emphasize the importance of communication in determining 'the things to which we attend' . . . ,"⁴ that is, in determining the extent to which decisions deviate from substantive (objective) rationality.

We need not agree that positive, neoclassical economics is effectively dismissed either by Simon's empiricism or Innis's historicism. The troublesome relationship between theory, history and empiricism in economics can be resolved in the context of decision theory, but not so easily or so destructively as both Simon and Innis seem to have thought; but that is another subject.

Keynes was a theorist, and the New Classical economists and Herbert Simon are theorists, though all are associated with methodologically distinct empiricisms. Innis used theory in historical analysis, thereby subjecting theory to history. New Classical, rational expectations theory presumes that decisions are based on unlimited reasoning power and unlimited, though costly, access to information. Accordingly, decisions are effectively maximizing. They are what Simon calls "substantively rational." Rational expectations theory admits to a degree of uncertainty in the information environment, but, by endowing agents with the ability to reduce uncertainties to true, objective probabilities, and by asserting that these probabilities are described by the statistical laws of random events, it can assert that any deviations from substantive rationality are minor and of short duration. Keynes, on the contrary, asserted that decisions are affected by habit, custom, historical factors, and psychological factors such as "herd instincts," that is, that decisions are substantively irrational. Irrational expectations determine actual outcomes.⁵ On this matter Keynes and the New Classical economists, obviously, do not agree. Both would agree, however, that the subject over which disagreement occurs is the degree of objective rationality in decisions or, in other words, the characteristics of the informational environment. Simon carried out psychological tests to specify the extent and the characteristics of the limits to rationality in decision-making. Innis's contribution was the description of historically specific limits to rationality. He described particular informational environments and he formulated the effects of specifiable conditions in those environments.

Over the past thirty years, Herbert Simon and his associates have provided the empirical foundation in psychology for Innis's reconstruction of the effects of information systems on the structure and evolution of institutions. Over the past fifteen years, debate between Keynesians and New Classical economists has re-emphasized the importance of the informational environment in the analysis of market and non-market behaviour. From the vantage point provided by these related research programs, it is possible to see Innis's work as a largely successful attempt both to present the evidence of history concerning the importance of informational environments and to formulate and corroborate hypotheses concerning the character of substantive non-rationality, that is, of the "bias" of decision-making behaviour.

INNIS IN THE NEW PERSPECTIVE

The multiplicity of commentaries since Neill's comprehensive study has revealed new details of Innis's position in the Canadian social science community. These details do not, however, constitute the basis for a re-interpretation. Rather, they help to fill out the picture which itself takes on meaning only when seen in the new perspective of theories of decision-making and of the informational environment.

In company with Adam Shortt, Oscar Skelton, W.A. Mackintosh, and a number of non-Canadian-born scholars, such as James Mavor and CR. Fay, Innis brought Canadians to an understanding of the key economic factors in the development of their country.⁶ He did so at a time when the academic disciplines of history and social science were being indigenized in Canada. Innis, especially, exploited the economic interpretation of history to make the point that a unique economy, built on a particular geographical and technological base, had given Canada a unique identity. There have been suggestions that Innis, insisting that foreign-trained scholars did not know the country, consciously exploited this view to improve the position of native Canadians in the academic community.⁷ Perhaps he himself rode the tide of Canadianization to what, it is alleged, he most wanted to have, advancement in the University of Toronto. Seen in this suggested perspective, however, Innis's behaviour follows logically from his conviction that the substance of Simon's conclusions was correct. He wanted immigrant economists to undergo a reorientation program lest they teach courses and prescribe policies based on inappropriate perceptions and preconceptions.⁸

Despite the wealth of new detail provided in recent years, commentators have continued to find difficulty in specifying a theory or view that can confidently be called Innis's scientific contribution, that is, some empirically tested or testable hypothesis or hypotheses of his upon which others might build. They have instead referred to his "concerns" or to an "Innis tradition."⁹ Despite his voluminous scholarly output, his own generation thought his work to be unfinished. There was talk, at first, that his son, Donald, or Marshall McLuhan might finish it. No one, except perhaps McLuhan, identified anything that could be called "Innis Law" or "the Innis Theorem," or even an "Innis Effect." W.T. Easterbrook's explicit pursuit of Innisian "concerns" in his elaboration of the historical effects of changes in the "macroeconomic uncertainty environment of decisions"¹⁰ was innocent of the more recent theories of decision-making, in the light of which, as I contend, Innis's contribution appears complete and his work finished.

It is now uncontroversial to attribute the Staple Theory of Canadian economic development to W.A. Mackintosh" rather than to Innis. The theory is too consonant with normative neoclassical analysis to be the product of a devoted disciple of Thorstein Veblen or of someone whose work can be identified with Simonean decision theory.¹² In fact, the idea that the Canadian economy had acquired a special character and set of policies from its reliance on a narrow, primary products export base was being applied by Adam Shortt¹³ long before either Innis or Mackintosh received his first degree. That the neoclassical version of the Staple Theory is now questioned even in the standard texts cannot, then, be taken as a direct criticism of Innis. In the 1970s a western Marxian version of the Staple Theory was developed with much reference to Innis, but the most capable exponent of the resulting "Theory of the Staple Trap" has clearly indicated that it cannot be drawn directly from Innis.¹⁴ Thus, Innis's contribution, in this regard, was not to establish Canada's reliance on staple exports; rather, it was to establish that this reliance was a less than efficient consequence of a distinctive, national informational environment.

It came as no surprise to Innis's contemporaries that a leftist theory, like the Theory of the Staple Trap, could not be drawn from his interpretation of Canadian history. Founding members of the League for Social Reconstruction, a sort of "brains trust" for Canada's socialist party, recall Innis as a pre-converted Paul acting on behalf of the "conservative establishment."¹⁵ There was no subsequent conversion in his case. Indeed, the picture that has been presented is not altogether appealing.¹⁶ While praising his political skill and his energy, Innis's contemporaries note how self-serving his purpose was as he rode the fashions of materialism, nationalism, and conservatism to the top of his profession. They recall that his disorganized and confusing presentations were somehow a success because they were at once methodologically fashionable and comforting to established interests. They accuse him of lacking a liberal education, of being intellectually naive, of being a perennial sophomore, forever half discovering everything. These personal reminiscences have special interest because Innis rejected the Keynesian paradigm that most of his contemporaries accepted.¹⁷ When put in perspective, they are enlightening with respect to Innis's contribution because he was as judgemental about the older, neoclassical paradigm as he was about its contemporary critics. It may be, as commentators seem to suggest, that Innis arrogated to himself the role of spokesman for scientific purity when, in fact, he was no purer than the others. What is of scientific interest, however, is the possibility of locating some objective meaning that informed his behaviour.

Innis received his intellectual formation during the early decades of the twentieth century when attention turned from moral to materialistic determinants of social behaviour.¹⁸ This new emphasis characterized the writing of Canadian history in the years of Innis's maturity. William Wallace, Frank Underhill, and Donald Creighton all drew extensively from economic interpretations of American history to tell Canada's story. Like Innis,¹⁹ they were not philosophically committed materialists, but they did let their attention dwell on the materialistic factors in historical development. A.R.M. Lower approached the problem of Canadian nationalism through the timber trade. Creighton approached Canadian imperialism through the commercial possibilities of a river basin.²⁰ Innis approached the rise and decline of empires about the North Atlantic through the fur trade, the fish trade, and the mining frontier. Collectively, they produced the idea that fur and the canoe had first united the northern half of North America, that wheat and the railway had reunited it, and that the federal government had been the political instrument of that reunification. Perhaps Innis was most responsible for this vision, but something very much like it had appeared in the writings of Edward Porritt and Gustavus Myers, as well as in certain regional complaints against the bias of federal power.²¹ The difference was that these voices sought to expose the inequities and inefficiencies of Canadian development. After Innis had turned his attention away from Canadian history, other, explicitly neoclassical scholars approached Canadian development as an optimal response to well-understood opportunities.²² Innis's position was, then, distinguished because it seemed acceptable in part to both views. Canadian development had been a limited success.²³ The result was not substantively optimal. Still, Innis concluded, the country was viable. We might say it was a typical piece of Simonean "satisficing."

Although their careers overlapped in time, there is no evidence that Harold Innis and H.A. Simon were in any way aware of one another. The similarities in their research programs was a consequence of their common intellectual roots in American Institutionalism. Innis turned to history, Simon to empirical testing. Their research programs, however, complement and support one another, each putting the other in a different perspective.²⁴ When Innis's work is seen in the perspective of decision-making theory, subsequent rejection of the normative, neoclassical Staple Theory of Canadian development²⁵ appears as a justification of his emphasis on the role of the information environment in producing substantively non-optimal decisions. In the light of Simon's critique of methodologies that combine quantitative analysis with normative neoclassical theory, the apparent contradiction between, on the one hand, Innis's repeated downgrading of studies that he considered "non-scientific," and, on the other, his rejection of "statistics" and of positivism disappears.²⁶ Innis rejected contemporary methods of positivism. He rejected the abuse of statistics in normative studies. He was not opposed to empirical science, to quantification, or, indeed, to the goals of positivism itself.

INNIS'S CONTRIBUTION IN PERSPECTIVE

In the second and third chapters of Reason in Human Affairs, Herbert Simon speculates about the evolution of institutions. His discussion is based entirely on empirical tests of his "behavioral model of rationality." Nowhere does he use the history of institutions to illustrate or corroborate his model; neither does he use the model to explain long-run historical phenomena, though that suggests itself as a natural extension of his inquiry. Innis, whose interests ran in another direction, undertook that extension. Innis's focus was narrow, returning persistently to the effects of the physical characteristics of information systems on attention, perception, and the consequent organization of decisions. He made frequent comments on such things as the efficiency of techniques of warfare, the effects of mathematics on computational capacity, the roles of emotion, of multiple goals, and of ignorance and uncertainty in the rise and decline of institutional arrangements; but he always returned to stress the physical characteristics of information systems, that is, to what he called the bias of communications. His distinctive contribution was to suggest testable hypotheses with respect to this influence on the long-run evolution of decision-making organizations.

Innis described the historical path of less than substantively rational decisions. He described the forces that he thought important in controlling deviations from objective rationality and in shaping the consequent organization of social behaviour. In normative, neoclassical theory, equilibrium is a fixed, conclusive termination of a substantively rational process, and that is all. Whether equilibrium occurs in particular historical conditions is a secondary concern, as is specification of the historical conditions under which the processes by which equilibrium is reached terminate. Innis recorded conditions that had caused deviations from substantive rationality and had frustrated the termination process. He called it the study of "economic pathology."²⁷ This research project was the common root connecting what has been called "the early" and the "later" Innis.²⁸ During the period when neo-Keynesian doctrines were the stuff of normal science in economics, when some basic elements in Keynes's analysis were pushed out of sight, Innis's preoccupation with these matters left him out of the mainstream of his discipline. However, with the advent of Rational Expectations theory and its insights into the conditions of objectively rational decisions, Innis's interest in the bias generating imperfections of the informational environment has regained relevance.

Commentators who have pointed to a relationship between the views of Innis and of the philosopher, George Grant,²⁹ have caught a glimpse of the Innisian project, but not of its definitive core elements. Innis did not share Grant's judgement that liberalism was morally vacuous. He was not a "Red Tory." He was an individualist, and a liberal in that sense, first, last, and always.³⁰ What Innis and Grant have in common is a perception of an informational environment in which it is possible for decisions to deviate from what is appropriate for institutional survival, that is, in which things can go wrong. Innis was not a philosopher. Paul Feyerabend subsequently developed a philosophy of science with which Innis would have been perfectly comfortable.³¹ Like Simon and Innis, Feyerabend focuses on the limitations of reason in human affairs. Feyerabend's "methodological anarchy" has the same relation to Innis's work as Simon's empirical decision theory. It is a research program with a set of conclusions that complements and supports Innis's historical analysis.

The Feyerabendian perspective on Innis (or the Innisian perspective on Feyerabend) may need clarification. Consider Feyerabend's repeated assertion that there are no objective facts, that all facts are theory-laden. Consider also his argument that we cannot separate the subject of science from its observational instruments,³² and immediately becomes evident that he was dealing with the bias of the media of communication, or the imperfections of the informational environment, however one wants to phrase it. Innis's reconstruction of the history of institutional arrangements turns on two factors. One, materialistic and Veblenesque, was the technical limits of the information media. In explaining the structure of decision-making in terms of available information and information-handling techniques, Innis reveals this Simonean element in his analysis.³³ The other was a moral or philosophical factor that Innis drew from his contemporaries in the Department of Classics at the University of Toronto, the interplay of reason and passion, of virtue and fate, in the informational environment.³⁴ While there are elements of this other awareness in Simon's work,³⁵ it is Feyerabend's understanding of the mere instrumentality of reason and its false pretentions to intellectual sovereignty that most closely suggests Innis's insight in this regard.³⁶

INNIS'S SCIENTIFIC CONTRIBUTION

Innis's contribution was the formulation and preliminary historical corroboration of several "operationally viable" hypotheses about the role of the informational environment in the structure of decisions.³⁷ Despite his often turgid exposition and his lack of a normal scientific presentation, his repetitiveness and his tendency to build up masses of examples make it relatively easy to extract four of the more important of these hypotheses. They should not be characterized as "laws." There may be no "laws" in the science of the artificial. They are better viewed as "effects." That is, they can all be characterized by a shared formula: given certain definable characteristics in the informational environment, certain definable behaviour patterns (i.e., institutional arrangements or structures of decisions) have been and will be a consequence. These effects may be called the Accelerating Disequilibrium Effect, the Initial Disorientation Effect, the Reorientation Conflict Effect, and the Minerva's Owl Effect.

The Accelerating Disequilibrium Effect is a generalization drawn from Canadian historical development. It may be stated as follows: when there is a social pre-commitment in uncertainty to large expenditures on physical and institutional capital, the resulting structure of alternative, subsequent decisions does not include an objectively rational decision to liquidate. That is, some decisions include a structuring of the informational environment that precludes reversibility. There are decisions from which there is no going back.

Innis owed the Accelerating Disequilibrium Effect to J.M. Clark. Like Innis, Clark attempted to extend Veblen's "Non Euclidian Economics." In the spirit of Simon's concept of bounded rationality, both Clark and Innis set out to make economics scientific by including in its considerations the constraints of engineering technique.³⁸ Clark's work was the foundation of the modern "Accelerator Theory" of business cycles, the idea that, when any other sector experiences growth, the capital goods sector accelerates the pace of growth by responding at an even faster rate. Clark originally developed the idea to explain alternating booms and slumps in frontier towns. Seeing the effect as a recurring phenomenon in Canadian history, Innis took it a step further by using it to explain the frontier's apparent inability to accept a strategic, organized reversal of expansion.

Innis turned the accelerator phenomenon into a general account of the effects of "overhead costs" in shaping the Canadian economy.³⁹ Development in a modern industrial economy required relatively large, up-front investments.

Arrangements for the placement of the moneys and provision for the long period between investment and ultimate pay-off required considerable overhead in the form of social organization. The consequence was possible heavy losses, especially for the decision-making elite that would have to authorize rational liquidation in case of failure. In narrowly based frontier economies, because region-specific and industry-specific capital could not be reallocated, the prospects for associated labour were only slightly better. Faced with their own possible bankruptcy, the decision-making elite chose a less than objectively rational alternative: they reinforced the original decision to expand. Innis did not assert, like Simon, that unit monetary losses are more heavily weighted in preference functions than are unit monetary gains; he did not question the scientific basis of subjective, expected utility analysis. Innis as historian simply noted that "the economics of losses is not less significant than the economics of profits."⁴⁰

In normative neoclassical economics it is assumed that market forces, in the long run, eliminate inefficient enterprise. Innis's point was that in certain circumstances, clearly documentable in the Canadian experience, something less than substantive efficiency is viable. Where there are industry- and region-specific activities involving significant private and social capital, and especially where there is a region-specific government with sufficient power,⁴¹ the information environment will generate recurring decisions to repeat strategies that reinforce inefficiency. He did not think the Canadian case was unique. The Accelerating Disequilibrium Effect seemed to him to be evident in the experiences of the United States and most European countries in the late nineteenth and early twentieth centuries.⁴² It is not difficult for anyone familiar with Canadian experience over the past ten years to think of evidence that would further corroborate the Effect. The histories of Dome Petroleum and the east coast fish processors are telling examples.

The Initial Disorientation and Reorientation Conflict Effects occur when the information environment undergoes fundamental restructuring as a result of a change in perception. According to Simon, who has established the empirical validity of the phenomenon,⁴³ the perceived identity of an object is dependent on the search procedures used in the recognition process. Search procedures, in turn, are dictated by the preconceptions of the searcher and by the sequence in which information is presented to him. Innis's point was that different media for presenting information are associated with different sequences of presentation and, consequently, different perceptions of objective reality. Preconceptions can be explained, at least in part, by the limitations of the various media used in earlier search exercises. These are the simple, basic elements of Innis's concept of the bias of an information environment. When fully elaborated, the concept is a very general notion, relating to certain specific media in much the way that Thomas Kuhn's notion of a scientific paradigm relates to the specific theory at its centre, by including the totality of the organization of the scientific community around the theory.⁴⁴

Initially, however, Innis considered only the technical characteristics of the media in question. A change in media characteristics constitutes an external, unexpected disruption of the informational environment. Rational expectations become impossible as the informational basis of expectations suddenly reforms itself through a radical change in perception. The consequence is uncertainty, disorientation, confusion, and paralysis in decision-making processes. This is the Initial Disorientation Effect. In time, a new informational environment is built into new structures of decision-making in which a new priority of values obtains. Conflict then occurs between new organizations developing under the influence of the new media, and older, established institutions.

References to the Initial Disorientation Effect are scattered through a number of Innis's essays. Consider, for example, the three following observations:

Instability of public opinion . . . follows the introduction of new inventions in communication.

Profound disturbances in Egyptian civilization . . . coincided with a shift in emphasis on stone as a medium of communication .. . to an emphasis on papyrus.

The consequent maladjustments were evident in the boom of the Twenties and the depression, and were to an important extent a result of expansion of the press and of a new instrument of communication—the radio. Public opinion became less stable and instability became a prime weakness, serving as a forced draft in the expansion of the twenties, and exposed to collapse in the depression.⁴⁵

It is best, however, not to multiply citations from Innis to show that there are testable hypotheses to be drawn from his work. His own presentation did not follow the canons and style of positive science, and his massing of overlapping examples tends often to confuse, rather than to focus, attention. For example, passages that describe the perceptual disorientation and instability caused by the advent of a new media of communication tend to run over into passages describing subsequent interest group conflict. Note, for example, Innis's statement that "The spread of writing contributed to the downfall of the Republic (of Rome) and the emergence of Empire."⁴⁶

Following an initial change in communication technique, institutions, according to Innis, organize around the consequent new perception of reality to exploit its possibilities. This is an "operationally viable," or positivist, hypothesis for which he provides ample historical support. The development of mathematics in Babylonia led to the reorganization of business enterprise. The development of a non-pictoral alphabet in Ancient Greece led to the reorganization of city states into the Hellenic empire.

Organization of decision-making and the ranking of priorities varied with the use of papyrus, parchment, and the printing press.⁴⁷ In every case, except those in which perception and organization were dominated by primitive oral communication, some form of inter-group rivalry was a factor in the result. Innis was convinced by his reading of history that organizations retaining the imprint of oral communication had a superior ability to absorb new values without conflict. In the absence of an oral tradition, technological progress in communication had been and would continue to be accompanied by wasteful conflict in which the more efficient alternatives might not be the winners.⁴⁸

In the context of empirical decision theory, the Reorientation Conflict Effect suggests experimentation with individual and group problem solving using different and changing means of communication. Innis did not take that approach, and no one else has tested his hypotheses. What Innis repeatedly did call for was some kind of balanced approach to the resolution of conflict between extremist views. He sought a greater degree of objectivity and adaptability in facing the perceptual consequences of changes in communication techniques, but he said nothing of what specific steps could be taken to meet this need.

Herbert Simon projected a "science of design,"⁴⁹ the subject of which would be the effects of changes in the bounds of reason on the organization of decisions. This was a first step in the direction suggested by Innis. Although he did not attempt it, Simon considered extending this "science of the artificial" to nations and multinational entities, that is, to the macroeconomic informational environment. By documenting a number of historical effects, Innis provided a beginning for the pursuit of Simon's project. The pieces fit. Simon's contribution led to a neoclassical elaboration of the economics of uncertainty and information,⁵⁰ the subject matter of Innis's work, though in a different paradigm.

Consideration of Innis's Minerva's Owl Effect⁵¹ leads in the same direction. The Effect suggests that the elements of an organization may survive a change in communication technique, but also that they may not, and that those most necessary to the conservation and advancement of the organization may not be the survivors. More specifically, the Minerva's Owl Effect suggests the petrification of a creative contribution to the ends of an organization as the contribution passes from one medium of communication to another. What was vital and growing in one informational environment, rigidifies in another, monopolizing thought and preventing further growth along the same lines. What had been common and useful, rather than sacred and revered, flowers into classic form and ceases to develop. It becomes a dead thing, disfunctional from the point of view of its earlier existence. As Innis noted in this regard, "Freedom of the press as guaranteed by the Bill of Rights in the United States has become the great bulwark of monopolies of time."⁵²

The Minerva's Owl Effect was, for Innis, the most personally disturbing of his hypotheses. In its light, he could see that the values and institutions that constituted the survival equipment of western civilization had appeared in different forms in technically different circumstances. These values and institutions had survived, but only under constant threat. As successive informational environments replaced one another, different elements in the basic values of western civilization blossomed in new forms and spread over the areas influenced by the relevant means of communication. Whether western civilization or, indeed, any civilization could survive these increasingly frequent jolts of technological change remained uncertain. Witnessing the disturbances of the Great Depression of the 1930s and of the subsequent total world war immediately after, it is not surprising that Innis was deeply pessimistic.

The Minerva's Owl Effect is related to the Reorientation Conflict Effect and, as such, could be readily tested in and refined in the context of empirical decision theory. A small group instructed in some written form to perform a co-operative task could be compared, with respect to its organizational behaviour, with another small group instructed by tape recorder to perform the same task. Alternatively, the same group could be instructed through different media to perform the same or a different task, with internal conflict being the object of observation. In this light, it is now difficult to understand why the media buffs surrounding McLuhan did not undertake such simple, positivist procedures.

THE REASSESSMENT

Innis's insights into the nature and importance of macro informational environments constitute a major contribution. He specified a set of testable, informational environment effects. His attempt to validate these hypotheses by historical reconstruction was a success. Of course, he made mistakes in detail. He tended to overgeneralize. He occasionally thought his generalizations to be validated in very dubious cases. However, quite apart from his position as a "conservative" at the beginning of a "liberal" era, and without reference to his position in the social science community in Canada, he made a recognizable scientific contribution on which others may build. Indeed, practitioners of the science of administrative design and, to some extent, those developing the economics of information, are proceeding toward the end product designated by him. Aspects of the "New Institutional Economics" in the Coase-Williamson tradition that are closely related to Herbert Simon's work⁵³ are also related to elements of Innis's thought. Still, no one has taken up the specific task toward which Innis directed attention, that of designing organizational procedures that would routinize adaptation to changes in media of communication.

On another level, in the new perspective in which Innis now appears, two additional tasks have become evident. First, with respect to social science in general, there is the need to work out a methodology for the science of design that can include history, empirical decision theory, and neoclassical economic analysis in mutually supporting roles, thus reducing the conflict that grows out of different disciplinary perceptions. Secondly, with respect to Canadian economic history, the elaboration of Canadian development as a series of satisficing decisions under the influence of biased informational environments, a study begun by Innis, has yet to be completed.

NOTES

¹ R.F. Neill, A New Theory of Value: The Canadian Economics of H.A. Innis (Toronto: University of Toronto Press, 1972).

² The work of H.A. Simon is well known. Nonetheless, I make special reference to his work on the significance of paths of information search (see Models of Thought [New Haven: Yale University Press, 1975], pp. 355-62) and of preoriented attention ("Rationality in Psychology and Business," Journal of Business 59 [1986], pp. 209-24).

³ Neill, A New Theory, p. 40.

⁴ H.A. Innis, The Bias of Communication (Toronto: University of Toronto Press, 1951), p. vii.

⁵ J.M. Keynes, The General Theory of Employment Interest and Money (London: Macmillan, 1936), Chapter 5, pp. 95-97, 148, 156, 203, 238.

⁶ Neill, A New Theory, pp. 35-40; C. Berger, The Writing of Canadian History (Toronto: Oxford, 1976), p. 84; C.D.W. Goodwin, Canadian Economic Thought (Durham: Duke University Press, 1961), pp. 176-95.

⁷ D. Creighton, "Harold Innis—An Appraisal," in W.H. Melody, L. Salter and P. Heyer, eds., Culture, Communication and Dependency (Norwood: ABLEX Publishing, 1981), pp. 13-25; E. Havelock, "Harold Innis: A Man of His Times, The Philosophic Historian," E.T.C. 38 (1981), pp. 141-68.

⁸ Neill, A New Theory, pp. 37, 65.

⁹ W.T. Easterbrook, "Innis and Economics," Canadian Journal of Economics and Political Science 19 (1953), pp. 291-303; M. Watkins, "The Innis Tradition in Canadian Political Economy," Canadian Journal of Political and Social Theory 6 (1982), pp. 12-33.

¹⁰ W.T. Easterbrook, "The North American Pattern of Growth and Development—The Continental Context," ed. Ian Parker (unpublished, 1984).

¹¹ M. Watkins, "The Staple Theory Revisited," Journal of Canadian Studies 12, 5 (1977), pp. 83-95; W.A. Mackintosh, "Economic Factors in Canadian History," Canadian Historical Review 4 (1923), pp. 12-23.

¹² Neill, A New Theory, p. 37; H.A. Simon, Reason In Human Affairs (Stanford: California University Press, 1983), pp. 12-17.

¹³ A. Shortt, "The History of Canadian Currency, Banking and Exchange: Lord Sydenham's Measures," Journal of the Canadian Bankers' Association 10 (1902), pp. 21-40.

¹⁴ D. Drache, "Harold Innis and Canadian Capitalist Development," Canadian Journal of Social and Political Theory 6 (1982), pp. 35-39; Watkins, "Staple Theory Revisited."

¹⁵ Havelock, "Harold Innis."

¹⁶ Berger, pp. 85-111; S.D. Clark, "The Contribution of H.A. Innis to Canadian Scholarship," in Melody et al. pp. 27-35; Creighton, "Harold Innis"; G.S. Dunbar, "Harold Innis and Canadian Geography," Canadian Geography 29 (1985), pp. 159-64; Havelock, "Harold Innis"; and A.R.M. Lower, "Harold Innis as I Remember Him," Journal of Canadian Studies 20, 4 (1985-86), pp. 3-11.

¹⁷ H.A. Innis, "Sub-Specie Temporis," Canadian Journal of Economics and Political Science 17 (1951), pp. 553-57.

¹⁸ S.E.D. Shortt, Search for an Ideal (Toronto: University of Toronto Press, 1976); Berger, p. 100.

¹⁹ Berger, pp. 55, 211; Neill, A New Theory, pp. 101, 17, 87.

²⁰ Berger, pp. 113-36; 236.

²¹ A. Rotstein, "The Alchemy of Fur and Wheat," Journal of Canadian Studies 12, 5 (1977), pp. 6-31; G. Myers, The History of Canadian Wealth (Chicago: Charles H. Kerr, 1914); E. Porritt, Sixty Years of Protection in Canada (London: Macmillan, 1908); C. Martin, The Natural Resources Question: The Historical Basis of Provincial Claims (Winnipeg: King's Printer, 1920); E.R. Forbes, The Maritime Rights Movement 1919-1927 (Montreal: McGill-Queen's University Press, 1977).

²² A.W. Currie, Canadian Economic Development (Toronto: Nelson, 1943); D.C. North, "Location Theory and Regional Economic Growth," Journal of Political Economy 63 (1955), pp. 243-58; R.E. Caves, "Vent for Surplus Models of Trade and Growth," in J.D. Theberge, ed., The Economics of Trade and Development (New York: Wiley, 1968), pp. 211-29.

²³ Neill, A New Theory, pp. 66-67.

²⁴Ibid., pp. 24-34, 40; H.A. Simon, "Rational Decision Making in Business Organizations," American Economic Review 69 (1979), pp. 493-513, 499.

²⁵ K.J. Buckley, "The Role of Staples Industries in Canadian Economic Development," Journal of Economic History 19 (1985), pp. 439-52; P. George, "Rates of Return for Railway Investment and the Implications for Subsidization of the C.P.R.," Canadian Journal of Economics 1 (1968), pp. 740-62; E.J. Chambers and G.W. Bertram, "Urbanization and Manufacturing in Canada 1870-1890," Canadian Political Science Association Conference on Statistics (1964), pp. 225-58. Also, E.J. Chambers and D.J. Gordon, "Primary Products and Economic Growth: An Empirical Measurement," Journal of Political Economy 74 (1966), pp. 315-52; R.E. Caves, "Export-Led Growth and the New Economic History," in J.N. Bhagwati, ed., Trade, Balance of Payments and Growth (Amsterdam: North Holland, 1971), pp. 403-42.

²⁶ H.A. Simon, Models of Bounded Rationality (Cambridge: Yale University Press, 1982), pp. 503-06; Neill, A New Theory, p. 41; W. Westfall, "The Ambivalent Verdict: Harold Innis and Canadian History," in Melody et ai, pp. 37-51.

²⁷ Neill, A New Theory, pp. 64, 148.

²⁸ G. Patterson, "Harold Innis and the Writing of History," Canadian Literature 83 (1979), pp. 118-30; H.A. Innis, Changing Concepts of Time (Toronto: University of Toronto Press, 1952), p. iii.

²⁹ A. Kroker, Technology and the Canadian Mind: Innis, Grant and McLuhan (New York: St. Martin's Press, 1984); John Hutcheson, "Harold Innis and the Unity and Diversity of Canadian Federation," Journal of Canadian Studies 17, 4 (1982-83), pp. 57-73.

³⁰ Neill, A New Theory, p. 10.

³¹ R.F. Neill, "Rethinking Canadian Political Economy," Studies in Political Economy 9 (1982), pp. 147-54.

³² P.K. Feyerabend, Against Method: Outline of an Anarchistic Theory of Knowledge (London: N.L.B., 1975).

³³ Neill, "Rethinking," p. 97; Simon, Models of Bounded Rationality, pp. 56-57.

³⁴ W. Christian, "Harold Innis as Economist and Moralist," Occasional Papers #2, University of Guelph, 1981; R. Keast, "It Is Written—But I Say Unto You': Innis on Religion," Journal of Canadian Studies 20, 4 (1985-86), pp. 12-25; A.J. Watson, "Harold Innis and Classical Scholarship," Journal of Canadian Studies 12, 5 (1977), pp. 45-61.

³⁵ Simon, "Rational Decision Making," pp. 29-38.

³⁶ P.K. Feyerabend, "Science, the Myth and its Role in Society," Inquiry 18 (1975), pp. 167-81; H.A. Innis, "Economic Nationalism," Papers and Proceedings of the Canadian Political Science Association for 1934 (1934), pp. 17-31; Innis, "Economics for Demos," University of Toronto Quarterly 3 (1934), pp. 389-95; Innis, "The Role of Intelligence: Some Further Notes," Canadian Journal of Economics and Political Science 1 (1935), pp. 280-87.

³⁷ P. Samuelson, Foundations of Economic Analysis (Cambridge: Harvard University Press, 1975).

³⁸ Neill, A New Theory, pp. 32-33.

³⁹ I.M. Spry, "Overhead Costs, Rigidities of Production Capacity and the Price System," in Melody et al., pp. 155-66; H.A. Innis, Essays in Canadian Economic History (Toronto: University of Toronto Press, 1956), pp. 141-65.

⁴⁰ Simon, Reason in Human Affairs, pp. 12-17; Simon, "Rational Decision Making," pp. 493-513, 506; Innis, Essays in Canadian Economic History, p. 272.

⁴¹ Neill, A New Theory, pp. 66-70, 146-49; H.A. Innis, "Government Ownership in Canada," Schriften des Vereins fur Sozial-politick 176 (1931), pp. 241-79.

⁴² H.A. Innis, "The Decline in the Efficiency of Instruments Essential in Equilibrium," American Economic Review 42 (1953), pp. 16-22; Innis, Essays in Canadian Economic History, p. 78.

⁴³ Simon, Models of Thought, pp. 292-305, 353-62.

⁴⁴ T.S. Kuhn, The Structure of Scientific Revolutions (Chicago: University of Chicago Press, 1970).

⁴⁵ H.A. Innis, Empire and Communications (Clarendon: Oxford University Press, 1950), pp. iv, 17; Innis, Changing Concepts, pp. 102-03.

⁴⁶ Innis, Empire and Communications, p. 121.

⁴⁷Ibid., pp. 40, 98-99, 140ff, 173ff.

⁴⁸Ibid., p. 41; Innis, Bias of Communication, pp. 92-93.

⁴⁹ H.A. Simon, The Science of the Artificial (Cambridge: M.I.T. Press, 1969).

⁵⁰ K. Arrow, The Economics of Information. Vol. 3: Collected Papers of Kenneth Arrow (Cambridge: Belknap Press of Harvard University Press, 1984).

⁵¹ Innis, Bias of Communication, pp. 3-52.

⁵² Innis, Changing Concepts, p. 108.

⁵³ Simon, Models of Bounded Rationality, pp. 480, 487.