I was born in the year 1911. The half century since that time has contained some of the most tremendous events in the history of the human race. Among them have been two world wars of unprecedented extent and violence, the near doubling of the total population of the world,³the rise of the great Communist states, and other events profoundly changing the course of human history. But preeminent among all are the growth of science and the growth of man's capacity to apply his mind to the problems of learning and discovery.

Many striking examples might be offered by the changes within science, and the changes wrought by science, in these 50 years. To me, a most striking illustration is a comparison of our knowledge of the universe in 1911 and now. In 1911 what men knew of space was confined to our own galaxy. Our solar system was thought to be near the center of that galaxy, whose shape was only dimly conjectured. Now we know that the sun and its secondary satellite, the earth, are far out on one arm of our vast, beautiful, spiral galaxy. We know also that there are at least a billion such galaxies within the space our telescopes have penetrated. Furthermore, we have seen the photographic record of objects five billion light-years away, moving away from us at half the speed of light. These 50 years have brought a more profound change in our knowledge of the cosmos than was achieved in all man's previous existence.

Although astronomy may stretch our minds most of all, there are other examples of advances in our learning and knowledge, of the deepest meaning and most comprehensive impact. Among them we might mention the general theory of relativity, the dismembering of the atom by nuclear physics, the discovery of the biochemical basis of heredity, the developments in engineering that made possible the Manhattan Project and the orbiting of men in space, and the chemical discoveries and developments in social organization that have promoted the world population explosion. It has been a truly epochal period, without any equal in history. Not least has been the final world acceptance of science as a tremendous social force.

As one views this panorama of glorious scientific achievement in the last 50 years, he cannot fail to be impressed by two things: the unity of scientific effort as it progresses; and great differences in the rates of progress among the subdivisions of science.

The first observation was skillfully described by the biologist Frank R. Lillie in 1915. "Scientific discovery is a truly epigenetic process in which the germs of thought develop in the total environment of knowledge. Investigation of particular problems cannot be accelerated beyond well-defined limits; progress in each depends on the movement of the whole of science."⁴

Lillie's observation must be considered in the light of the second point, the differentials of progress among separate subfields of the scientific community. The progress of "science as a whole" at any given time in large measure is the progress of a relatively few subjects with growing points. As growing-point salients move, they furnish ground for practitioners in other disciplines to stand on and in turn push into new territory. This is what makes intercommunication among the sciences so important, and even more the proper choice of those with whom we communicate. To paraphrase an ancient observation, every scientist stands on the shoulders of giants. But one might add that it is important to stand on the shoulders of the right giant. The selection is as important as the standing. In the period between 1910 and the mid-1940's, physics and the mathematical disciplines stood out as examples of the giants.⁵ Chemistry was of shorter stature in this comparison, biology considerably shorter, and geology less visible. Comparisons among the social sciences were more difficult, but perhaps anthropology, psychology, and economics deserve some distinction for their accomplishments in the pre-World War II period. However, the differences among subgroups within a field were in most cases as pronounced as differences between major fields.
 
 

We are, naturally, especially interested in the place that geography occupied in this advancing front of science. There is no reason to avoid frankness. I am sure that all but a few here would agree that our contributions have been modest thus far. We have not been on the forward salients in science, nor, until recently, have we been associated closely with those who have. The reasons are not difficult to find. During the early part of this 50-year period, in the 'teens and early twenties, our closest associations were with history and geology. Geological study of that period, and of the thirties, was not among the inspiring growing points in science. The history and the geology connections did not correct the predisposition of our scholars of the 'teens and early twenties to the deceptive simplicity of geographic determinism. This was perhaps one of the last appearances of the Newtonian view of the world.

As determinism began to fade and independent geography departments sporadically appeared in this country, geography turned to association with the social sciences of the period. "Possibilism" in man's relation to the earth took the place of determinism. Because of the limitations of the social sciences and history at the time, these associations were only slightly more productive sources of inspiration than geology. It was only much later, indeed in the early fifties, when association with the social sciences bore its soundest fruits for geography. This was in the methods descended from mathematical statistics, first applied in biometrics, anthropometry, and econometrics. Their full application has not yet run its course.
 
 

I began my professional interest in geography at a time when the old moorings to geology were almost severed. The groping for solid footing among the social sciences was well under way.⁶ The geographers who turned in the direction of the social sciences made a prescient choice of direction, but the difficulties confronting us were enormous, considering methods then at our disposal. In the face of those difficulties, it was only natural that we became somewhat introspective. We tried to build a platform, as it were, from our own materials and to anchor it ourselves.⁷ This search for a professional identity was, of course, found during other periods in the history of geography. It goes back at least to the 19th century German geographers. Alfred Hettner and others in Germany undertook influential studies from the early 1900's onward. But the succession of methodological appraisals in the United States that commenced with Harlan Barrows' "Geography as Human Ecology"⁸ in 1923 and continued for nearly forty years must certainly rank as one of the most intensive efforts toward this end.

Our search for a professional identity led to an intellectual independence and eventually to a degree of isolation against which a number of the rising younger generation of geographers have now reacted. In our search for solid footing, a meaningful image of ourselves, many of us tended to separate ourselves from other sciences. Our principal interdisciplinary communications were with other sciences which also had problems of isolation, like cultural anthropology and geomorphology. In effect, some of us saw geography as an end in itself rather than in the broader context as a contributor to a larger scientific goal. Perhaps this is the fate of many specializations.

Insistence on the independence and separation in the 1930's and 1940's may seem a shockingly incorrect statement to some of you. Did not geography alone recognize its relations to both the physical sciences and the social sciences? Did not geography deal constantly with the data accumulated through the efforts of other disciplines? Indeed, was not geography even alert to analogous methods of inquiry from other disciplines? One can cite such major statements in the field as Barrows' "Geography as Human Ecology," and Sauer's "The Morphology of Landscape"⁹ as proof of this alertness. But I must note that both these statements came in the mid-twenties, and thereafter for at least 25 years an atmosphere of separatism and independence characterized the profession.¹⁰ Furthermore, morphology was not a particularly happy choice as an analogue method, and the hint given by Barrows on ecology was never seriously followed up by his colleagues. For science at large, morphology already was becoming a somewhat sterile concept when we took to it, and the analytical methods of the twenties and thirties were not yet equal to the multivariate problems of ecology.¹¹ The concept that became dominant among us was that of "areal differentiation," derived from Hettner and introduced in the United States by Sauer.¹² This concept favored (although did not demand logically) a goal of investigation independent of the goals of other sciences. The same might be said of another important concept in the field, that of areal functional organization, introduced by Platt.¹³ On the other hand, the work of Sauer and his disciples did find common ground with cultural anthropology, but it also was a somewhat isolated science until the 1940's.

In our desire to make our declaration of independence viable, we neglected to maintain a view of the advancing front of science as a whole. We acted as though we did not believe in anything more than the broadest generalities about the universality of scientific method. In effect we neglected to appraise continuously the most profound current of change in our time. We neglected an axiom: The course of science as a whole determines the progress of its parts, in their greater or lesser degrees.
 
 

What did we miss in the course we took? For one thing, we missed early contact with developments in mathematical statistics, and early touch with antecedents of systems analysis. The scholars whose thought influenced life (and social) systems concepts greatly, like R. A. Fisher and Karl Pearson in biology and anthropology, Alfred Lotka in biology, Sewall Wright in genetics, and L. L. Thurstone in psychology, were all active in the 1920's and 1930's. The flowering of the application of their techniques and concepts awaited the availability of electronic computers and mathematical progress in the late 1940's and early 1950's, but they provided forceful organizing advances in genetics and other biological fields, in physical anthropology, demography, psychology, and economics from 15 to 25 years earlier than in geography. We thus missed for a period the new thought their techniques generated, because the techniques were essential keys to communication of that thought.¹⁴

Within the last decade we have made good our initial failure to respond to these modern techniques. We have even felt the influence of physics, as a few have experimented with the application of physical analogues to the phenomena of distribution. Although not a few among us have been uneasy about their meaning, these techniques already have proved their power. Mathematical analysis is a recognized part of instruction in alert departments of geography. We can only welcome the growth of these methods, because then have been a notable and needed stimulus to the rigor of our thinking. Even more important, they increase our capacity to communicate precisely with workers in other fields of science.

Is the mathematicization of our discipline the way of our future? In a sense, yes. The year is not far off when a geographer will be unable to keep abreast of his field without training in mathematics. Furthermore, he will find it increasingly difficult to conduct meaningful research without such training. But here we must enter more than a word of caution. There is a great deal more to science than the application of mathematics, or of rigorous logic. We must take care to examine carefully the paths of research down which our computerized mathematical colleagues lead us, or perhaps push us. The danger of a dead end and nonsense is not removed by "hardware" and symbolic logic. Before we go too far we should see what else there is about science at large that produces its "growing points." What determines how productive the use of statistics and hardware will be? In a few other fields scientists are facing problems of this kind that are somewhat out of control today. Recent attention to the scientific part of space exploration is an illustration.¹⁵
 
 

Can we make any observation about the methods of science at large that will enable us to keep needed mathematicization under control? There are a great many definitions of science. I am sure that many of you are familiar with most of them. One definition I like is: "Science is a quest for regularity underlying diverse events." This quest proceeds through careful, verifiable observation and description; through the construction of hypotheses, to project reality into the unknown; testing of the hypotheses through the conduct of experiment or further observation;¹⁶ replication of experiment and observation; and the building of a body of theory from verified hypotheses which in turn becomes the basis for new hypotheses, and new observations and experiments. Mathematical and statistical analyses have found their important place in this procedure because they aid in obtaining exact observation, and because they aid enormously in designing hypotheses that lead into the unknown.

I might stop here, and you would recognize this as a portrait of science. However, it is a portrait only of its skeleton. Three important additions provide the all-important life and direction that have figured wherever great strides in science have been made. I have already mentioned one: cross-disciplinary communication.

The second is what some men have described as the intuitive side of science. Warren Weaver has said, "...science is, at its core, a creative activity of the human mind which depends upon luck, hunch, insight, intuition, imagination, taste, and faith, just as do all the pursuits of the poet, musician, painter, essayist, or philosopher."¹⁷

But there is more to it than this. The mind of the scientist, no less than that of the poet or musician, must be structured by thought and experience before it reaches the creative stage. Some persons are able so to structure their minds more easily than others. It has been said, for example, that Irving Langmuir always saw matter, of whatever form, wherever he was, in terms of its molecular structure thus opening the way automatically for his many remarkable insights. Every scientist does this in some degree. There is no doubt that there is such a thing as "thinking geographically." To structure his mind in terms of spatial distributions^l8 and their correlations is a most important tool for anyone following our discipline. The more the better. If there is any really meaningful distinction among scientists, it is in this mental structuring. It is one reason why we should approach the imposition of analogues from other fields, as from physics, with the utmost care. The mental substrate for inspiration does differ from field to field.

A third important ingredient of science is highly developed sense of problem. In my pleasant and valued association with Professor Charles Colby at the University of Chicago, I can remember his frequent reference to the cultivation of such a sense. I now realize how wise and perceptive his advice was. In my duties of the past five years at the Carnegie Institution of Washington, I have had to maintain current knowledge about research in several biological and physical sciences. In all of them this sense of problem is very keen where outstanding progress is being made. Herbert Simon has observed that science is essentially problem solving.¹⁹ This observation is so important that it deserves a few words of elaboration. A sense of problem, at its most meaningful, is really a sense of the hierarchy of problems in a broad field, and possibly all science.

Every major field with which I am familiar has an easily recognized overriding problem. The overriding problems always lie behind the frontiers of investigation. They are remarkably few, and all fade into infinity in their ultimate forms. Indeed, the overriding problems of all science may be reduced to four: (1) the problem of the particulate structure of energy and matter, which physics treats; (2) the structure and content of the cosmos, which astronomy, astrophysics, and geophysics treat; (3) the problem of the origin and physiological unity of life forms; and (4) the functioning of systems that include multiple numbers of variables, especially life systems and social systems. Others might express these problems differently, but I believe that each of them is a beacon orienting research on the frontiers of the rapidly advancing fields.

Beneath each overriding problem are major second-level problems, and finally the problems translatable directly into experiment or observational investigation. For example, a major secondary problem related to the overriding one of the origin of life is the description of life in pre-Cambrian times. It is translated directly into a search in pre-Cambrian rocks for stable chemical compounds known to be indicators of life. In this way Philip Abelson and his collaborators at the Carnegie Institution have produced firm evidence of the existence of life at least 2.6 billion years ago.

The same relation among the hierarchy of problems can be seen in growing-point research in astronomy, geophysics, biology, and elsewhere. I do not mean that all research is so organized, or is distinguished by the sense of problem. Most commonly an appreciation of the hierarchy of problems is shared by relatively few in each field. It is indeed one of the most troublesome questions facing the administrator of public research funds in the nation at the present time.
 
 

By now my theme should be obvious: The geographer should seek his personal identity in the mirror provided by all sciences. How is this translated into future geographic progress? The development of a professional identity in geography has two aspects: the future development of the theoretical study of spatial distributions; and a reappraisal of the overriding problem recognized by our discipline.

The first, further development of the theoretical, is our true inner refuge as specialists. It is what helps to structure the mind "geographically." The more rigorously the structuring is done, the more likely the discipline will have a cutting edge that places it on a research frontier. However unrelated and esoteric it may seem, the cultivation of theoretical study of spatial distributions is basic.

If we have had any generally accepted overriding problem in the past, it is areal differentiation, a concept widely accepted and usefully employed, particularly by American and German geographers. Its rationale has been ably presented and skillfully defended by Richard Hartshorne. His most recent definition of areal differentiation as the "accurate, orderly, and rational description and interpretation of the variable character of the earth surface"²⁰ still stands as a useful general guide to geographic method. A second preoccupation, but less widely held, was with the geographical expression of culture processes. I shall refer mainly to areal differentiation in the remarks of the next few paragraphs.
 
 

At a time when the social sciences provided us with very little firm assistance, and we were stressing our independence, areal differentiation of the earth's surface did serve as an overriding problem. It is time that we recognize the limitations of this concept. Do we need something more for a purposeful selection of research problems leading us to significant research frontiers? If we look at the concept of areal differentiation carefully, we see that it did not often lead us to common ground with the other sciences. We see it also as ending in a somewhat static goal. In effect, it stressed a hierarchy of regions as our hierarchy of problems.

I suggest that we take a fresh look at the hierarchy of problems, ignoring for the moment some of our traditional points of view. I noted earlier that science is problem solving. The problems that can be examined meaningfully depend on the methods which are available for their solution. As the centuries have gone on, men have steadily increased their capacity for problem solving, but the truly important changes in methods of problem solving have been remarkably few. They might read somewhat as follows: writing; Arabic numerals; analytical geometry and calculus; and the combination of techniques that comprise systems analysis. There was a time, perhaps just after the Second World War, when the inclusion of systems analysis in such a list might have been considered controversial. That is no longer true. Systems, as you know, are among the most pervasive and characteristic phenomena in nature. Each human being, man or woman, is a system, that is, a dynamic structure of interacting, interdependent parts.²¹ Perhaps that is less appealing than a poet's definition of a pretty girl, but it has meaning in that it relates the girl as a system to all other systems, such as a colony of ants, or a city, or a business corporation.

Systems analysis provides methods of problem solving which might be said to have been created for geography, if there were not also many other uses for them. Geography is concerned with systems. Indeed, we may now state its overriding problem. It is nothing less than an understanding of the vast, interacting system comprising all humanity and its natural environment on the surface of the earth. This might be compared with Humboldt's statement of a century ago, "Even though the complete goal is unobtainable,...the striving toward a comprehension of world phenomena remains the highest and eternal purpose of all research."²² It may also be compared with Hartshorne's definition of the purpose of geography as "the study that seeks to provide scientific description of the earth as the world of man."²³Compare also Barrows' "geographers...define their subject as dealing solely with the mutual relations between man and his natural environment. By 'natural environment' they of course mean the combined physical and biological environments... Thus defined, geography is the science of human ecology."²⁴All these statements have some similarity. However, the concept of the world of man as a vast interacting, interdependent entity permits us an effective orientation to a set of problems at different levels in a way that we have never had before.²⁵ Furthermore, it puts us in a context of sharp new problem-solving methods.²⁶ If we are willing, it also places us in association and in close communication with other sciences whose overriding problems are similar.

Viewed in this way one can see a host of beneficial results. We no longer are concerned about whether what we are doing is geography or not; we are concerned instead with what we contribute toward a larger goal, however infinite it may seem. As in other sciences, an overriding problem of infinite extent should be a challenge, not cause for resignation or despair. We no longer debate about whether geography can construct "laws." At the same time we do retain an identity by structuring our minds to handle spatial distribution patterns in all their complexity. But as we go about our task of analyzing spatial distributions and space relations on the earth we should keep in mind the question, "What, if anything, do geographic observations and analyses tell us about systems generally, and the man-environment system particularly?"
 
 

We might elaborate this position in summary manner: (1) The basic organizing concept of geography has three dimensions. They are: extent, density, and successions.²⁷ "Spatial distribution and space relations" are a verbal shorthand for describing the dimensions of the concept. A theoretical framework for investigation may be developed from this basic concept, as observations confirm hypotheses.²⁸ (2) The universe treated by geographers is the worldwide man-natural environment system. Geographers share their overriding problem, an understanding of this system, with other sciences. (3) The worldwide system is composed of a number of subsystems. The subsystems assist in identifying a hierarchy of problems for research. (4) The techniques of systems analysis are of particular value to geographers in applying their organizing (space) concept to the analysis of subsystems of the worldwide man-environment system. These techniques, because of their rigor, permit replications of analysis and comparability of results among different research investigations. They also state the results of geographic research in terms comparable to those of other sciences using systems techniques, and therefore make such results of greater potential use in treating the overriding problem, or any subproblem.
 
 

Events in the world of today make it absolutely essential that geographers adopt such a view if they have aspirations to the frontier of research. Not only do much sharper probes exist for examining man's activities, but society itself is responding to scientific change. It is being organized in ways that are more easily evaluated. The scientific revolution we have been going through is being accompanied by a revolution of rationalism in our economic structure. Indeed, it has been called a "second Industrial Revolution," with effects already very profound for all humankind. Industrial engineering years ago removed the individual decision making of the artisan. "Cybernation," or systems design and systems engineering,²⁹ are now rapidly moving individuality from "middle management" decision. This development is part of the social problem of automation. Not least, systems design and engineering, through the nation's defense programs, is having a dominant role in domestic political affairs and international relations. Research approaches have even been made toward understanding the process of human thought itself. Herbert Simon has said, "We shall be able to specify exactly what it is that a man has to learn about a particular subject--...how he has to proceed--in order to solve effectively problems that relate to that subject."³⁰ And, as you know, already a great deal is known about manipulating some aspects of society, like consumer demands, in a more or less controlled fashion. What we in the United States are experiencing is also going on in Europe and in Japan. Quite a different form is found in the Soviet Union, but it still is certainly an aspect of rationalization. We may expect similar developments in other parts of the world. And we may expect systems engineering to play an increasingly large role in coping with the social and economic crises that technological change has brought.³¹

These events and trends have the profoundest significance for the future spatial distribution of human activities, and we could not hope to anticipate or understand that distribution without being fully abreast of what is taking place. On the other hand, there must be something that the study of spatial distributions can tell us about these phenomena. To say this in brief, the methods that have created important salients on the frontier of the physical sciences are changing society itself, both directly and through their impact on the behavioral sciences.
 
 

We are, then, concerned not only with a vast interacting system, but with one that is being altered by knowledge of systems. We now come to the most difficult part of our determination. Recognition of the overriding problem is of little significance unless we relate it to the direction of everyday research, and, by extension, to the fields with which we seek common ground in the definition of problems. What does this tell us about our own frontiers?

The one thing that most distinguishes a system is the flow of information within it. "Information" is not to be confused with the ordinary meaning of the word, for it refers here to any mechanism that holds together the interdependent, interacting parts of a system. This is an interesting and critical point as far as geography is concerned, because the connectivity within a system is its most important characteristic. Many geographers, on the other hand, have stressed differences, as exemplified in the term "areal differentiation." If you accept my proposal of the overriding problem for our science, it then follows that to choose a research problem without reference to the connectivity of the system is to risk triviality. What space relations tell us of connectivity in the system is significant to science as a whole. Areal differences are significant only insofar as they help to describe and define the connectivity or "information" flow. We now see that the geographers who have been concerned with cultural and other processes have had an insight of significant direction in research. Eight such processes were suggested in the past--four physical and four cultural. Among the cultural you may remember demographic movement, organizational evolution, the resource-converting techniques, and the space-adjusting techniques. Among the physical, dynamics of the soil mantle, movement of water, climate, and biotic processes were suggested.³²

A second important characteristic of a system is the existence of subsystems within it. The pretty girl, if you like, can be broken down into an astonishing number of subsystems, like any complex being. The same is true of other complex systems. This is another important and critical point, for we must make the proper selection of subsystems for study if we are to maintain significance. We already have a clue in the past suggestions made about the importance of processes. It is the functional subsystems that are generally the significant ones. Thus the systematic aspects of geography, insofar as they treat functions, are disposed to a high level of significance. Those geographers who have thought in terms of areal functional organization again have had a significant insight as to research direction.

However, not all types of region have equal significance for research. Political regions are territorial units with a high level of significance because they are functional. A watershed is an example of a physically determined region that is significant. On the other hand, the old concept of a "geographic" region may have very little significance. We may need to review critically the significance of other types of regions within the context we are considering. The concept of a region is potentially valuable in systems study, but we should take care that the regional concepts we actually use are significant to the overriding system.
 
 

This brings us through the second level in the hierarchy of problems, down to a level where one must seek specific examples of significance. As geographers have long appreciated, the flow of "information" within the man-natural environment system is indeed vast. Selection of a research problem at random again risks triviality, even though it may be entirely "geographic" in conception. At this point one commences to be most actively concerned about clues from other sciences as to significant working problems.

Here we may go back to one of the first observations made in this discussion: The sciences differ enormously in their rates of progress. For example, not all divisions of the behavioral sciences or the earth sciences offer channels for productive communication. Without doubt we can benefit greatly from some collaborative definition of research problems with other sciences, but the cooperation must be selectively chosen. A good rule of thumb would be: Where systems analysis techniques are understood and incorporated at the working face of the discipline, a collaborative definition of problem may profitably be sought. In other words, cooperation is likely to be rewarding where methods made familiar in the physical sciences are now reaching into the neighboring earth sciences and the behavioral sciences. Where the concepts and approaches using systems analysis methods are making inroads, a possible place of interest is suggested for geography. Relations with other sciences which at times have been loose, vague, and hard to define may thus become more meaningful.

The profession is becoming equipped gradually to take such a view in its fundamental research. The wind of change which we have felt for the last decade includes the application of some methods of systems analysis. Thus far they generally have been the application of more rigorous techniques to old geographic problems. Except for collaboration with economists and others of the "regional science" group, and the older collaboration between cultural geographers and cultural anthropologists, we thus far have done relatively little to explore common ground with other sciences on the definition of significant problems. In almost any direction we turn, interesting possibilities appear. Indeed, there are so many opportunities that the number of people undertaking geographic research seems remarkably few.

The relation of geography and the neighboring natural sciences is particularly interesting. By the neighboring natural sciences I mean studies that focus on the surface features of the earth, like soils, biotic features, and water movement. The logical point of contact of these sciences with the human part of the great man-land system is geography. In all of them there is increasing appreciation of the role of man. For example, it is realized that pollution has become a major feature in world hydrology; biological ecologists now admit that even the most "inviolate" natural preserves will be affected by man, no matter what protection is given; and a few geomorphologists now recognize the significance of man as a part of geomorphic processes. We should he particularly alert to overtures from these neighboring sciences, like that of Geoffrey Robinson in geomorphology, who suggests that at least some geomorphologists are interested in a collaborative definition of problems.³³ We should continue to capitalize on a point of view that geography alone, until recently, has maintained among the sciences concerned with man: land is half of the man-land system.

There are signs that geography's position as a "gateway" between the behavioral sciences and the earth sciences is being challenged somewhat by the behavioral sciences themselves. Economists, for example, in the last ten years have become increasingly concerned with natural resource development problems. To be sure, geographers helped to start them along these lines, but there is now a direct working relation between economics and hydrology. It is significant that the aspects of economics emphasizing a systems approach provided the important recent contributions to study of resources.
 
 

These events suggest that we need to maintain a comprehensive view of the frontiers in the behavioral sciences, and that we have a good clue to common interest in looking for those investigators who pursue a systems approach. It has been said, "The behavioral sciences are diverse in subject matter and state of development, yet ideas and concepts circulate quite freely among them...."³⁴ The quotation may be a slight overstatement, but it does represent an agreed-upon ideal in these sciences which we might well contemplate. How far do we join them in the shaping of goals and in the exchange of methods which have commenced to use?

An illustration or two may direct our attention to possibilities. We have mentioned that a most important characteristic of a system is the flow of "information," broadly defined. "Information" may be in the form of goods, people, messages containing data or ideas, or other dynamic phenomena. The geographer, by definition, looks at what spatial distributions tell concerning this information flow, or vice versa. Geographers already have attacked some of these problems of information flow successfully.³⁵ Probably the most important general question of this kind familiar to geographers is: "What can we say about how people distribute themselves and their culture on the earth, given free choice?" Much of the work geographers have done thus far is within the context of economic constraints, but they also respond to their concepts of amenities, to neighborhood and other group attachments, to the diffusion of information, and perhaps to other factors. There is a wealth of significant problems here to examine. Attention to them can bring us into a common area with students of motivation in the behavioral sciences. This is a key area in behavioral science research. We may find eventually some interesting common ground with psychology, thus finally connecting with the inferences of M. G. Kendall twenty-five years ago.³⁶ Indeed, study of the brain is considered one of the most useful approaches to the study of systems generally.³⁷

Geographers recently have been alert to noneconomic "information" flow studies. For example, a much respected pattern for geographical research with mathematical methods has been the diffusion of innovation studies by Torsten Hägerstrand in Sweden.³⁸ These studies, well known to American geographers, have stimulated diffusion research in this country. Such research is also a natural outgrowth of long-continued American interest in diffusion phenomena, followed particularly by cultural geographers.³⁹

At the same time the interest in diffusion studies illustrates our past relations with other scientific subjects. American sociologists have been carrying on very similar work since the early 1940's, including some elaborately designed experiments.⁴⁰ As far as I can discover there was little cross-disciplinary communication on this remarkably similar path of research until about a year and a half ago. It is obvious that collaboration here between geography and sociology can be of value. This is of more than academic interest, for as Ullman has noted, "the relative 'stickiness' of society, the resistance of certain areas to spread of innovations and improvements," has strong implications for public policy both nationally and internationally.⁴¹

Allow me another and more unusual example. An interesting offshoot in the behavioral sciences at the present time is the study of conflict theory, to which Kenneth Boulding of the University of Michigan and others have contributed. Looked at from the point of view of systems and their information flow, conflict theory is essentially a search for "redundancy," or the capacity to handle in channels multiple movements with the same destination. Boulding has suggested that the theory may be of interest in studying land use.⁴² Here is an opportunity to help in exploring the overriding system through a fresh idea.

A common front with the behavioral sciences is important not only in framing significant research questions but also because of geography's long association with historical study. Increasingly, it looks as though history would acquire scientific meaning through the dimensions given it by behavioral science.
 
 

We emerge with four general points that could help to place our science on a research frontier. (1) Continue to strengthen quantifying methods. Attempt to add to them rigorous analytical approaches in our theory and habits of constructing hypotheses. (2) Recognize an earth-wide man-environment system as our overriding problem. We can seek significant research questions in the study of subsystems at different levels, amenable to our spatial distribution analyses. (3) Choose our research problems in the light of the advancing frontier of the behavioral sciences, and with attention to systems-oriented study in the neighboring earth sciences. Finally, (4) supplement our present heavy commitments to studies within economic constraints and to morphology studies by other approaches. The rising interest in cultural geography is healthy, but we could diversify still more. I particularly commend to your attention political geography within the systems framework. It is concerned with regions that have true functional significance in the great man-land system.

Seeking and staying on a research frontier is a most exacting task. It is now very clear that, in this age of specialization, special knowledge and specialized concepts are not sufficient to hold a science on the frontier. The sense of overriding problem is essential, and so is a view of at least a part of the spectrum of all science. This does not mean that future accomplishment will be entirely by those who are mathematically sophisticated. For those of us not so endowed it is comforting to remember that A. A. Michelson, the first American to win the Nobel Prize, was, by his own admission, poorly prepared in mathematics. But he did have an extremely keen sense of the overriding problem in his field, a passion for exactness, and an alertness to the contributions of neighboring disciplines. There is an important place for a comprehensive view, but it must be a view based on something more than undergraduate and graduate courses. I believe the time is near when post-graduate training and a second doctoral degree may be the price for reaching a research frontier. In our plans for future professional action and in our advice to those in professional training, we must think about these matters before it is much later. If we do not, others will cultivate our frontier, for that is the way of science. If we do, perhaps we may come closer to justifying Charles Darwin's words, "...that grand subject, that almost keystone of the laws of creation, Geographical Distribution."⁴³

1. Address given by the Honorary President of the Association of American Geographers at its 59th Annual Meeting, Denver, Colorado, September 4, 1963.

2. This paper makes no pretense to coverage of all the ways in which geography may be viewed, or practiced. It discusses geography as a science. There is equal justification for placing some geographic scholarship among the humanities, as William L. Thomas has noted in a letter to me (June 24, 1963). If such a distinction is needed, one might follow Howard Mumford Jones in his definition: "The humanities are...a group of subjects devoted to the study of man as a being other than a biological product and different from a social or sociological entity" (Howard Mumford Jones, "What Are the Humanities?," One Great Society [1959], p. 17). Insofar as we encounter spatial distribution entities not amenable to the methods of science, and of interest to any serious scholar, our subject does have a humanistic content. But one may also question, as some scientists do, the appropriateness of these dividing lines between science and the humanities. As Marston Bates provocatively has said, "...science is only one of man's approaches to the understanding of the universe and of himself. By understanding,...I mean trying to make sense out of the apparent chaos of the outer world in terms of the symbol systems of the human mind. This might be considered the function of all art; and in that case I am led, half seriously, to call science the characteristic art form of Western civilization...the sciences and the humanities form a false dichotomy, because science is one of the humanities" (Marston Bates, "Summary Remarks: Process, "Man's Role in Changing the Face of the Earth" [1956], p. 1139). Compare also William Shockley, "...the practice of science is an art" (Science, Vol. 140 [1963], p. 384).

3. Estimated mid-1963 world populations: 3.25 billion; 1910 populations 1.7 billion (1963, extrapolated from United Nations data; 1910 extrapolations from estimates by W. F. Willcox and A. M. Carr-Saunders).

4. Frank R. Little, "The History of the Fertilization Problem, Science, Vol. 43 (1916), pp. 39-53.

5. The above comparison is not intended to reflect popular, or even professional, evaluations of the time. General appreciation of events in mathematics and physics during the late thirties, for example, did not come until the mid-forties. Yet they were sources of basic thought on methods that have affected all sciences.

6. Cf. J. M. Blaut, "Objective and Relationship," The Professional Geographer, Vol.14 (1962), pp. 1-7. "In this respect we behaved like the social sciences: our philosophical weakness, like theirs, had its roots in chronically unsolved problems. Their problems concerned values, causes, and social wholes. Our problem, then as now, concerned the nature of our subject matter."

7. The work of Carl Sauer and the "California School" in collaboration with cultural anthropology was an exception.

8. Harlan H. Barrows, "Geography as Human Ecology," Annals, Association of American Geographers, Vol. 13 (1923), pp. 1-14.

9. Carl O. Sauer, "The Morphology of Landscape," University of California Publications in Geography, Vol. 2 (1925), pp. 19-53.

10. The drive for the independent department typified this atmosphere at the time. Again the interest of the California group in cultural anthropology may be cited as an exception.

11. Barrows had true insight in stressing "place relations," but his concept of geography as human ecology set forth too ambitious a field. Neither qualitative nor quantitative methods of the time offered much solid ground for exploiting the ecological concept. At least in retrospect we can see the ecological concept of Barrows' time as incompletely formed (i.e., the adjustment of an organism to environment). It has now been replaced by the much more powerful monistic concept of an ecosystem, in which organism and environment are one interacting entity.

12. Sauer, op. cit., p. 20.

13. See R. S. Platt, Field Study in American Geography, University of Chicago Department of Geography Research Paper 61 (Chicago: 1959), especially pp. 302-51.

14. An interesting demonstration of these techniques falling on sterile ground in geography occurred in 1938, when the mathematical statistician M. G. Kendall presented his paper "The Geographical Distribution of Crop Productivity in England" before the Royal Statistical Society (Journal Royal Statistical Society, Vol. 102 [1939], pp. 21-62). This study was an analysis of covariance among ten crops in the 48 English counties. Besides the interesting direct conclusions he drew, Kendall made some provocative observations about the similarity of statistical techniques for studying a psychological problem and for studying a geographical problem. However, the two geographers present, L. Dudley Stamp and E. C. Willats, devoted their comments on Kendall's paper mainly to its shortcoming in interpreting the observable landscape. So far as I know, there was no sequel in geographical study to Kendall's interesting exploration. I am indebted to Brian L. Berry for calling my attention to Kendall's paper.

15. Some scientists fear that space "hardware" is causing an inefficient, even dangerous, misallocation of high-quality scientific talent in the United States. (See P. H. Abelson, Testimony before the United States Senate Committee on Aeronautical and Space Sciences hearings on National Goals in Space, June 10, 1963.)

16. The geographer may observe through field investigation; he may experiment with the use of statistical models (or idealized reality).

17. Warren Weaver, "Science, Learning and the Whole of Life," Address at 70th Anniversary Convocation, Drexel Institute of Technology, December, 1961.

18. By "spatial distributions," "earth-spatial distributions" is, of course, understood here. They are the parallel of distributional associations in other sciences.

19. Herbert Simon, The New Science of Management Decision (New York: 1960), p. 34. There are other similar statements, like that of T. S. Kuhn, who calls it "puzzle-solving" (The Structure of Scientific Revolutions [Chicago: 1962], pp. 35 ff.).

20. Richard Hartshorne, Perspective on the Nature of Geography (Chicago: 1959), p.21. The concept of areal differentiation as Hartshorne explains, "stems from Richtofen's synthesis of the views of Humboldt and Ritter and has been most fully expounded in Hettner's writings" (ibid., p. 12).

21. A useful short categorization of systems is given by Kenneth E. Boulding in his "General Systems Theory--The Skeleton of Science," Management Science, Vol. 2 (1956), pp. 197-208. He distinguishes nine "levels" of systems in increasing order of complexity. A social system is of the eighth order among his levels.

22. Alexander von Humboldt, Kosmos: Entwurf einer physischen Weltbeschreibung, Vol. 1 (Stuttgart: 1845), p. 68. Quoted from Richard Hartshorne, Perspective on the Nature of Geography, p. 162.

23. Hartshorne, Perspective on the Nature of Geography, p. 172.

24. Barrows, op. cit., p. 3.

25. The closest approach to this in the geography of the '30's and '40's was in Robert S. Platt's view of geography "as the science of regional process patterns of dynamic space relations." Platt, "A Review of Regional Geography," Annals, Association of American Geographers, Vol. 47 (1957), p. 190. However, the appropriateness of formal systems concepts to geographic research is not mentioned by Platt.

26. A very gracefully stated description of the indivisible attribute (and others) of systems is given by Sir Stafford Beer in "Below the Twilight Arch--A Mythology of Systems," in Systems: Research and Design (Donald P. Eckman, ed.) (Wiley, New York: 1961), pp. 1-25.

27. Extent is measurable as size, shape, and orientation. Density is shown by the amount of "betweenness." Simultaneity is a special case of succession.

28. Cf. Blaut (op. cit., pp. 5-6), who interprets Hartshorne (op. cit., pp. 74-80, 133, 144-45) and states the organizing concept as "areal integration." I do not find Blaut's statement inconsistent with the statement given in this paragraph, but his does leave the epistemological problem of what space is. (Discussed by Blaut in his "Space and Process," Professional Geographer, Vol. XIII [July, 1961], pp. 1-6.) In addition, the word "integration" has a connotation of study technique that (to me) detracts from clarity.

By extension I also do not find the statements of this paragraph inconsistent with Hartshorne's latest careful analysis of geographic concept and method (Hartshorne, op. cit.). It may be noted that Hartshorne, always precise in his definitions, has described the components of geographic study in a manner that allows them to fit the view of geography suggested here, and probably other views also.

29. See Donald N. Michael, Cybernation: The Silent Conquest, for a summary account of the social changes caused by systems engineering. Simon, op. cit.. also describes them.

30. Simon, op. cit., p. 34.

31. E. A. Johnson has stated one aspect of this problem, from a national point of view: "...the increase in physical knowledge has made the future...uncertain, ...we must plan much further ahead in a way that will provide much greater flexibility, whether this be in peaceful or military affairs, whether it be for the individual or for the country. ...our primary problem is to find a way to manage our very big systems affairs in this new situation. We will have to examine our individual, group, and national values to see what it is we want to do in a rapidly changing world, and to see what we can do consciously to manipulate in our favor the real and perhaps hostile physical and world environment so that it will serve us better. This is a problem of big systems." E. A. Johnson, "The Use of Operations Research in the Study of Very Large Systems," Systems: Research and Design (Donald P. Eckman, ed.), pp. 52-93.

32. Ackerman, Geography as a Fundamental Research Discipline (Chicago: 1958), p.28.

33. Geoffrey Robinson, "A Consideration of the Relations of Geomorphology and Geography," The Professional Geographer, Vol. XV (1963), pp. 13-17.

34. Behavioral Sciences Subpanel, President's Science Advisory Committee, Strengthening the Behavioral Sciences (Washington: April 20, 1962), p. 13.

35. A number of reports on such research have appeared in the Annals; e.g., articles by W. L. Garrison and others. Publications of the "regional science" group also are illustrative.

36. See footnote 14 above.

37. Beer, op. cit., p. 19. "The brain is itself the most resplendent system of them all..." We may well reflect to what degree social reality reflects the structure of the brain.

38. Torsten Hagerstrand, "The Propagation of Innovation Waves," Lund Studies in Geography (Sweden: Lund, 1952) and succeeding publications.

39. See, for example, Fred Kniffen, "The American Covered Bridge," Geographic Review (1951), p. 114.

40. See, for example, James S. Coleman, "The Diffusion of an Innovation among Physicians," Sociometry, Vol. 20 (1957); Melvin DeFleur and Otto Larsen, Flow of Information (New York: Harper, 1958); Anatol Rapoport, "Spread of Information through a Population with a Social Structure Bias," Bulletin of Mathematical Biophysics, Vol. 15 (1953).

41. Edward L. Ullman, "Geography Theory in Underdeveloped Areas," Essays on Geography and Economic Development (Norton S. Ginsburg, ed.), University of Chicago, Department of Geography Research Paper 62 (Chicago: 1960), pp. 26-32.

42. Kenneth Boulding, Conflict and Defense (New York: Harper, 1962), p. 1.

43. Charles Darwin, letter to Joseph Dalton Hooker. 1845.