Geographers are wont to boast of their subject as a very old one, extending, even as an organized science, far back to antiquity. But often when geographers in this country discuss the nature of their subject, whether in symposia or in published articles, one has the impression that geography was founded by a group of American scholars at the beginning of the twentieth century. Likewise many such discussions in the past have had only slight relation to similar discussions by European geographers, in spite of the well known fact that geography has had a greater development in the universities of Germany alone than in those of this country.

European geographers, notably the Germans, have been much more generally interested in the study of the nature of geography than those in this country. Many of the questions raised here only in recent years have been very effectively discussed by certain German geographers since the beginning of the century. It seems desirable therefore to examine these discussions at some length. If satisfactory answers can be found there to objections raised by our critics we may be spared the disturbance created by apparently new suggestions for radical departures from established lines of work, and those working in established lines may be assured of the value of their efforts. On the other hand if these many studies of the problem indicate that in certain respects we have been following false hopes, further loss of effort may be spared.

The studies of the nature of geography on which this paper is based are listed, together with illustrative studies, in the foregoing bibliography. Some effort was made to include in the former all of the more significant studies of recent years--though it is no doubt deficient in respect to the French literature, not to mention those in languages other than English, French and German. No such effort was made in connection with the illustrative literature however, the writer merely using whatever studies happened to come to his mind.

In a critical survey of this character, based almost entirely on the work of living colleagues in his field, a writer might prefer to omit specific references, but this easier choice is forbidden him because of his responsibility not only to the reader but also to those colleagues whose work has provided the material for the study. (See index of authors at end.) In particular, the critical position assumed places an imperative responsibility upon the writer to provide a means by which quotations, paraphrases, and critical comments may easily be checked against the writings concerned. Consequently a voluminous number of individual references are required. To avoid the disruption of the reader's thought created by so many footnotes, a simple reference system, suggested by the editor, is followed. The numbers in brackets provide the necessary basis for reference without, it is hoped, unduly interrupting the reading.

It would of course be absurd to suggest that American geographers were entirely unfamiliar with the concepts of geography developed in Europe. (Early in the nineteenth century Jedidiah Morse worked in close relationship with European geographers [409]). Arnold Guyot of Princeton, the first professor of geography in an American university, was a follower of Carl Ritter, but he had no successors. Far more important was the missionary work of Ellen C. Semple, whose presentation of Ratzel's concepts of anthropogeography set the course of human geography in this country for a quarter of a century. During the same period, William Morris Davis and Professor Albrecht Penck of the University of Berlin did more than any others before or since to establish mutual relations between geography in this country and Germany. Not only was there a close connection in their own work, but each taught for a year or so in the other country--Penck at Yale and Columbia, Davis at Berlin--and each published to some extent in the other language. Their field of mutual interest, however, was almost exclusively limited to physiography. The emphasis that Davis gave to this part of geography combined with Semple's exclusive consideration of human geography developed much the same form of dualism that was characteristic of an earlier period in Germany. Davis' attempt to bridge the gap was purely theoretical [203], as he himself made little effort to make the connections in practice.

For nearly two decades these two points of view dominated most of the thought of American geographers, both in their research and in their concepts of their subject as presented, particularly, in the presidential addresses of Brigham, Dryer, Barrows, and Whitbeck [305; 207; 208; 212]. The notable exception offered by Fenneman [206], influenced by the work of Hettner, appears to have made little impression until some years later. With that exception, geography was conceived of as a science of relationships between the natural environment and human activities. This "environmentalist concept" of geography was given its most careful statement by Barrows, who, reversing the usual form, speaks of geography, or "human ecology," as the study of man's adjustments to the natural environment.

A somewhat different point of view was introduced into this country from France by the publication in English translation of works of Vidal de la Blache and Jean Brunhes. Although both of these writers retained in theory the concept of geography as the study of relationships between man and the earth, in detail their work pointed in a different direction, and particularly tended to emphasize regional studies [Brunhes, 182, 4, 13-27, 552 ff., 83, 55; Vidal, 184, 3-24; cf. Sauer, 84, 171, 180-1].

The greatest change in point of view, however, has been caused by the work of Carl Sauer in bringing the attention of American geographers to concepts developed in Germany nearly a generation ago. Through his three somewhat similar studies, published between 1925 and 1931, as well as by his marked personal influence, he has led many, particularly among the younger workers, to think of geography in terms of the study of material landscape features, both natural and cultural, and to consider these features according to their chorographic, or regional, interrelations [211; 84; 85].

American geography has therefore been markedly influenced by the work of European geographers, but it is notable that this influence has come through but very few contacts--for the most part through Davis, Semple, and Sauer. A few individuals, like Fenneman, or Wellington Jones, have been independently influenced, but have had little effect on the concepts of geography held by the rest. To be sure, many American geographers have followed the work of British geographers, but unfortunately the latter have been as provincial as ourselves if indeed not more so. The exceptions appear conspicuous in their lack of influence. Thus Chisholm, in his inaugural address at Edinburgh in 1908 [192], abstracted at some length "the most comprehensive and to my mind the most illuminating recent series of papers on the scope of geography," namely Hettner's classic statement of 1905. But this view of the field, essentially, as we shall see, that of most German geographers, appears to be known to but few in England today. Similarly the views that Herbertson brought back from his studies in Germany made little impression in England until relatively recently. The great majority of English geographers appear completely unaware of the developments in German geography since Ratzel (note, for example, Bryan's frank admission of this fact in a footnote report of a conversation with one German geographer--whose name is misspelled and whose views are misstated [280, 7] ). Most English geographers, depending largely on Semple for their knowledge of Ratzel, have tended greatly to overemphasize his importance in Germany [cf. Roxby, 195, 280; and Dickinson and Howarth, 10, 195-202]. In more recent discussions, however, Dickinson has strongly emphasized the concepts of German geographers of this century [101; 202]; further, a number of English geographers have been influenced by the work of Passarge.

The situation in France has been somewhat similar. Vidal de la Blache and Vallaux gave great attention to Ratzel's work, but in the later methodological discussions of Vallaux and of Brunhes one finds little understanding of what German geographers since Ratzel have thought or achieved. Consequently the several publications in English of these French geographers have done nothing to lessen American ignorance of modern German geography.

In return, we may add, German geography has been but little affected by the methodological views developed in other countries. As exceptions, we may note the influence of the regional monographs produced under the direction of Vidal de la Blache, which undoubtedly helped to further the recent emphasis on regional geography in Germany; and Herbertson's efforts to establish "natural regions of the world," which for a time gained more attention in Germany than in England. The publication in German of lectures by Sir Francis Younghusband has influenced a number of present German geographers, notably Banse and Volz. Apparently the only American geographer who has notably influenced German geography to date was W. M. Davis, whose developmental system of land forms classification was taken up by Penck, and is also represented in one of the standard German texts of geomorphology of which Davis and Braun were co-authors. German students of this subject continue to argue over the relative merits of the "Davische" system. On the more general aspects of methodology of geography as a whole, however, Davis appears to have had little influence in Germany.

These relatively limited associations in the methodology of the subject are in marked contrast to the increasing degree to which American geographers are utilizing the research products of European geography. Koeppen has become almost a household word in our considerations of climatology; the work of Russian and German students has revolutionized our classification of soils; Siedlungsgeographie has added new words like Strassendorf to our terminology; Passarge's concepts of Landschaft find expression in at least one recent American text; those interested in political geography have been made familiar with the developments in this field in Europe [216]; and Joerg provided us with a survey of the whole field of European geography during the World War and post-War periods [88].

In spite of this development, however, the tradition apparently remains that any question as to the nature of the subject is a matter of personal opinion rather than one requiring study. Even the animated discussions that followed Sauer's challenging statements appear to have led few to examine the original studies on which his conclusions were based. With the exception of Hall's résumé of the concept of regions [290] and a short discussion by the writer [216, 795-804], most geographers have been content with second-hand quotations, or have depended on their own consideration as to what geography ought to be. The most important exception is Finch's recent presidential address, published as this paper was nearly completed [223].

It has often been suggested that the question of the nature of geography is of minor importance--that what is needed is simply for geographers to work. "Es ist viel herumgedoktert worden an der Geographie" Penck complains, and states that "progress in a science is not made by writing about methods, but by working methodically" [163, 50]. But we may presume that Penck, who has published at least eight methodological discussions, does not mean that it doesn't matter whether we know where we are going so long as we are on the move. In a field in which the members who are officially recognized as "geographers" may, at a single meeting, read papers which to some of the hearers appear to belong in geology, climatology, soil science, economics, history, or political science, it is particularly necessary that we know what our field is, in order that we may hope to understand what its individual students are trying to do.

Even more important is the need for each individual who proposes to devote his professional life to the field of geography to have a clear picture of the scope and nature of that field. Not that any prescribed limits should be outlined within which every one who is called a geographer must confine his work; no individual or group is given any authority in science to prescribe such limits nor to determine the character of the work to be carried on within them. But experience has all too clearly demonstrated that those who enter the field of geography with preconceived ideas of what it is or should be are afflicted with a feeling of dissatisfaction, if these ideas prove to be in conflict with what geography actually is. Whether this feeling of disatisfaction be directed at themselves or at their colleagues, it represents an unhappy waste of intellectual energy deleterious to the effective progress of both the individual and the field as a whole. It is therefore one of the primary duties of those who attract young students into any particular field to give them so accurate a picture of the nature of that field that any later disillusionment will be unnecessary. But we cannot do this unless we ourselves know the nature of our field.

No one can read or listen to many of the discussions on the nature of geography by American geographers during recent decades without observing a marked motivation of reform. Evidently there is a certain general dissatisfaction with the field of geography, at least in its present state. To some degree at least these symptoms are echoes of similar symptoms that may be observed in the literature of European geographers, notably of the Germans.

Needless to say, a feeling of dissatisfaction with the present state of any field of knowledge--as distinct from a dissatisfaction with its essential nature--is a symptom of healthy vitality. But violent or arbitrary reactions to that feeling are not evidences of maturity. A bright-eyed youngster may observe errors in the way in which the pieces of a stained-glass window have been put together, and if he demonstrates his point by aiming stones accurately at the points in question he will create a notable disturbance but he will have done little for the improvement of the window. In many such cases the following statement of Humboldt seems appropriate: "It is far from my practice to find fault with endeavors, in which I have not myself made any efforts, because their success so far appears very doubtful" [60, I, 68].

In discussing one of the more violent reformers among German geographers, Gradmann has asked: "Is it really completely unavoidable that our science should be exposed naked before all the world, in that every new year a new 'reformer' arises who condemns the whole previous works to the foundations and feels himself called upon to erect something entirely new out of the ruins" [251, review, 552 f.].

We have already suggested--and the later consideration of one or two of the more famous attempts to reform geography radically will demonstrate the suggestion--that a thorough dissatisfaction with the state of the field is a phenomenon whose explanation involves the critic as well as the object of his criticism. Although we are justified in keeping this possibility in mind there is no intention of dismissing in any such manner any proposals for reform in a field of study for which none of its students would claim perfection. Each of the important proposals of recent years will be analyzed on its own merits, in its relation to the historical development of geographic thought and in its relation to a logical concept of the field.

Several different tendencies may be observed in the various proposals that have been urged for the reform of geography. The current Weltanschauung in German thought finds expression both in the demand that geographers should give greater emphasis to the importance of individual men responsible for the historical development of the features of any region, and in the demand that geography must serve the national interest [e.g., Schrepfer, 174]. To anyone familiar with the German geographical literature of the World War and post-War periods the suggestion that German geographers have been deficient in studying geography in Germany's interest is at least as surprising as to the elder German geographers themselves. On the assumption that this movement is not likely to make a positive impresion in this country, we need not consider it further [the interested reader will find pertinent discussions in Hettner, 175, 341-3; and Plewe, 177, 226 ff.].

A second tendency, of greater importance in this country, seeks to add to geography some element that provides "problems." The "mere description" of the character of a region, even with the addition of interpretation, apparently, is not a problem, or at least not a problem in the sense in which that term is used in some other sciences. Consequently, it is thought necessary to add the element of change in time, to compare the situation of today with that of a previous time, to study the effects of some particular change in one factor in a region upon the others, or, in general, to make geography "dynamic" rather than "static."

The remaining efforts to reform geography that need to be considered appear to be much more definitely motivated by a desire to make it what apparently it has not been, a proper science. In the first place, since science must deal with observable phenomena, geography, in order to be a science, should limit itself to observable phenomena, and these it finds in the visible "landscape"--or, if it is claimed that some things are observable but not visible, geography must limit itself to material objects. These, in contrast with immaterial phenomena, we may observe and measure with some degree of accuracy and certainty.

Another suggestion is based on the assumption that each particular science has its own distinctive objects or phenomena to study. A science of geography must therefore somehow discover its distinctive objects of study. Most of the visible or material objects that it finds in areas have already been claimed, if not thoroughly exploited, as properties of other sciences. One possibility, therefore, is for the geographer to pick up whatever objects he can find that no one else has claimed, and mark these for his own.

Other geographers insist however that the proper objects of study for geography are the pieces of land which he calls regions. In order that these may be studied as other scientists study their objects, arranging them in classifications of types and developing scientific principles or laws about them, they must be concrete unitary objects. If one questions this assumption but accepts the major premise as to the nature of science, then the study of regions is not science but perhaps a form of art, which must be excluded from a "science of geography."

At this point one is reminded of a remark of Professor Whitehead when confronted by one of his students with a discontinuity in his logical reasoning: "What we need at this point is a beautiful phrase." The escape from the dilemma in regard to areas as objects is aided by one of those blessed words of double-meaning--"landscape." Since the landscape is the visible scene, including within it visible material objects, and may be limited in various undefined ways, it is called a unitary concrete object. But the landscape in its German form "Landschaft," if not in English--is also the limited area of land, province or region. Hence the region or landscape is a concrete unitary object having form and structure and therefore subject to classification and the development of scientific principles concerning its relations.

The detailed examination of the nature of geography which this paper endeavors to present is not based on any assumption that geography is or ought to be a science--or that it ought to be anything other than it is. Assuming only that geography is some kind of knowledge concerned with the earth, we will endeavor to discover exactly what kind of knowledge it is. Whether science or an art, or in what particular sense a science or an art, or both, are questions which we must face free of any value concepts of titles. We will, therefore, not ask ourselves whether geography, as Douglas Johnson puts it, can "expect to enjoy the high prestige of the sciences," whether it "should pretend to a position of equality among them," or should accept a more humble though "honorable place along with history, economics, and sociology." Such questions naturally produce only the reaction that Johnson himself expressed: a refusal to "surrender geography's claim to a place of equality among the sciences until the scientific possibilities of the subject are more thoroughly explored than they have as yet been" [103, 220]. Geography is thereby put on the defensive, ever attempting to be something that perhaps it cannot be.

On one aspect of this question we may clear the ground at the start. That geography could be "a science in the sense that geology and botany are sciences" is as impossible as it would be for botany to be a science in the sense that geology is a science, or vice versa, or for either to be a science in the sense that physics and chemistry are sciences. Whatever geography is, its venerable if not honored age would nullify the most enthusiastic efforts of any students to make it over into something entirely different. Geography is not an infant subject, born out of the womb of American geology a few decades ago, which each new generation of American students may change around at will. Granted that one must keep an open mind on any subject in science, we are hardly free, as one unnamed geographer claimed, to change our concept of geography each year as the spirit moves us [see quotations in Parkins, 105, 222].

In any established field, the problem of arriving at its proper definition is not so much a creative task, a problem in logical reasoning from a priori assumptions, as a research problem in knowledge itself. The philosopher, Kraft, in subjecting the field of geography to a critical examination--to which we will refer in detail later claims that the problem of the position of any subject within the totality of sciences is a work that extends beyond the particular scientific knowledge of the department concerned, is properly a problem "in scientific theory, i.e., in the theory of cognition." But the function of such a research in the theory of science cannot be to prescribe what geography should be, but rather to ascertain what it contains [166, 1, 3]. Similarly Hettner wrote, in 1905: "The system of the sciences has been an historical growth; abstract designations of the sciences that tend to take no account of the historical development--unfortunately the methodological literature of geography has been particularly rich in such a priori conceptions--are foredoomed to unfruitfulness" [126, 545].

This point of view has recently been characterized by Leighly as blind adherence to "tradition" in the face of "strict logic" [222]. Even if the concept of geography that he defends could be shown to be logically sound--a question to which we will come later--the argument is not to be won by devitalizing the historical evolution of a field of study as mere "tradition." When one considers the cultural origins of the inhabitants of the United States, one might claim that logically they should speak either a variety of European and African languages or a mixture in which each was represented in proportion to its part in the total population; it is merely a "tradition" of our schools that all should be taught something approximating "the King's English."

The growth of knowledge in any branch of science depends less on the particular scope for which it may find a logical justification, than on the continuity of its life, from the past through the present to the future. In the history of science--as in all history--common knowledge among educated people so emphasizes the achievements of a few outstanding "masters" as to lead one to suppose that each of them, starting from scratch, had singlehanded accomplished his wonders. In geography, we have a striking example in Alexander v. Humboldt. He is frequently presented as a pioneer without predecessors, a lone figure who explored America and returned to establish the science of physical geography. In reality, as we shall shortly observe, it my be questioned whether Humboldt would have become a geographer at all, had it not been for the younger Forster, and the latter's geographic work was very largely dependent on the elder Forster; his antecedents we do not know. Furthermore Humboldt throughout his work depended on a veritable host of other workers, both of earlier times and of his own time, as the multitudinous references in his writings indicate.

If, in contrast, each new generation of geographers, or each individual geographer presents new suggestions, untested either by example or by study of the previous development of the subject, there is introduced, as Hettner observed, a feeling of unrest and insecurity among geographers that handicaps the successful cultivation of the field and, in addition, causes workers in other sciences to become sceptical of the whole subject [152, 52 f.; 167, 265; 168, 490 f.]. American geographers are all too familiar with such results. If we are to continue to experience violent shifts of the helm--formerly toward physiography, then toward environmentalism, now toward landscape studies, tomorrow to "the topography of art" and thereafter who knows whither¹--our ship will beat around with ever-changing aim, hence aimlessly, and will arrive nowhere.

Throughout the preparation of this study the writer has assumed that few if any readers are interested in the particular view that he as a single geographer may have of the field. The privilege of presenting statements of that kind is reserved, one presumes, for the masters of our field whom this Association elects to its presidency. The writer's concern, on the contrary, is to present geography as other geographers see it--or have seen it in the past. Since the geographers of any particular time and country represent but a small part of the whole field, it will be necessary to extend our examination both into other countries and into the past. If we wish to keep on the track--or return to the proper track, as was suggested in a recent symposium--we must first look back of us to see in what direction that track has led. Our first task will be to learn what geography has been in its historical development.

Reversing the usual procedure, therefore, we will not start with any particular concept of the nature of science, but rather with the field of geography as it has been produced. Having ascertained its essential characteristics we will endeavor to find what logical position it holds among the different branches of knowledge [cf. Kraft, 66, 3]. The first of three major portions of the paper is therefore concerned with the question "What is geography?" (Secs. II-IV).

On the foundation thus established we will be in a position to examine various suggestions that have been made for improving geography. This examination of major problems in geography today constitutes approximately half of the total study (Secs. V-X). The reader will understand that the brief statement of suggestions given above (Sec. B) does not do justice, even in outline, to the ideas of the students proposing them. Because many of these students have made distinguished contributions to geographical research their theoretical suggestions for its improvement demand our careful attention. On the other hand we may presume that they themselves will wish that their ideas should not to be taken on authority but should be subjected to critical examination. Since in many cases the significance of a theoretical suggestion can most clearly be seen where it has been applied in practice, our examination may also require a critical survey of the material products. In both cases the reader is particularly requested to keep in mind that it is only the validity of the ideas of the nature of geography that concerns us here. The informed reader will recognize from the bibliography that I have confined myself largely to the writings of geographers of established distinction. Since the standing of these writers is in nearly every case based on their non-methodological works, it cannot be affected by any discussions in this paper. It is the methodological ideas of these writers that we are concerned to question: are these, either in terms of logic or of fact, valid statements of the nature of geography or acceptable suggestions for improvement within geography? If these questions cannot be answered in the positive, it will be necessary to consider a sound logical basis on which geography may develop as a broad but unified field.

Finally, by way of conclusion, we may seek to relate the field of geography, in its essential characteristics, to the general nature of knowledge, to determine, that is, what kind of a field of study geography is (Secs. XI-XII). Whether it is properly to be called a science or not, we will leave to those more interested in terminology; if we know definitely what kind of a study it is, we need not concern ourselves over titles.

The writer wishes particularly to stress, that in considering the suggestions of his colleagues, his purpose is not to engage in dialectic arguments, seizing any opportunity that lack of clarity of statements might offer, but rather to understand the intention of the suggestions made. In particular he has no wish to introduce into American geography a form of personal argumentation that has unhappily marked the methodological discussions of German geographers in recent times. With Obst, his purpose, is to seek for the maximum degree of understanding and agreement possible, not to establish any particular direction in geography but to make possible a common basis for geographic work, so that the geographic work itself--the only real purpose--may be furthered to the highest degree [178, 1-3].

In a brief examination of published studies on the history of geography, John K. Wright commented on the lack of any adequate treatment of this subject in English [9]. We have numerous excellent treatments of the history of American geography, notably those by W. M. Davis [102; 104], Douglas Johnson [103], and Colby [107]. Attention may also be called to the recent publication in German literature of two separate surveys of recent developments in American geography, by Broek, of the University of California, and by Pfeifer, who has also worked in the same institution [108; 109; but note Platt's critical comment, 224, 125]. The development of geography in this country, however, can only be understood in the light of earlier and contemporary developments in Europe; for a thorough treatment of this subject we still look in vain in our literature. Since Wright first noted this lack, in 1925, Dickinson and Howarth have published a useful survey of the field, but one that unfortunately depends largely on foreign interpretations of the work in Germany, and gives but the briefest sketch of the developments in that country since Ratzel [10].

There can be no question that the foundation of geography as a modern science was primarily the work of German students. In whatever country one starts, the study of the development of geography leads backward to the work of Humboldt and Ritter in the early part of the nineteenth century. Further, though in many countries geographic work has become more or less independent, major changes in geographic thought are found to be expresions, frequently very belated, of the developments in German geography. It is particularly unfortunate therefore that we have in English hardly even an outline of the development of geographic thought in Germany during the past century, so that nearly all American geographers have been trained in complete ignorance of the methodological background of their subject. German students, in contrast, have readily available not only the original writings of the period, but also many historical studies of the development of geographic thought.

To present this history adequately would require a book in itself. For our present purpose we must confine ourselves to the consideration of the major characteristics of geography as a field of study at each of several successive stages in its development. Our concern is not with the development of geographic work, nor with the relative importance of different writers or groups in the production of geographic material, but rather with the establishment and changes in thought concerning the nature of geography, as represented however in geographic works as well as in direct methodological studies.

Our examination of earlier work in geography has a two-fold purpose. We have already emphasized its importance in providing us with an understanding of the character of the field in which we work. The second purpose will be to observe what earlier writers have concluded in regard to problems which are of current concern in the methodology of geography, for there are few of these that have not claimed the attention of geographers in the past--in most cases, repeatedly in the past. As Braun observed, "there will be no improvement in the methodological confusion as long as we geographers continue the habit of completely ignoring the writings of our predecessors in this problem . . . and attempt to construct a building with stones that the neighbor has perhaps long ago recognized as useless or insufficient for a better arrangement" [155, 17 f.]. "Any progress in methodology," wrote Obst, "must be built on the past. No methodological change therefore without knowledge and consideration of past understandings and concepts" [178, 3].

These two purposes of our historical study cannot be attained if one enters the study with a pre-established concept of what geography should be, and on these terms searches among the writers of the past for those whose ideas are in agreement with ours, whom one then places in opposition to those who appear to have taken a different view of geography. In any methodological debate, to be sure, one may look for supporting arguments wherever he may find them--whether in the present or the past--but such a presentation is not to be confused with an historical study which aims at an objective presentation of the past; on the contrary it leads easily to a distortion of the past development of geography and of the ideas of earlier geographers in order to fit the purposes of the particular debate. We will therefore attempt to consider the concepts of previous geographers, so far as is humanly possible, independent of our own particular views.