I call these observations 'crucial problems' in memory of Roman Jakobson, who from time to time in the early 1960s used to present a course at MIT entitled 'Crucial Problems in Linguistics,' which a number of us attended and much appreciated. What defined Jakobson's course was the fact that the subject matter was open-ended. On the first day of class he would solicit suggestions from us students as to what problems we would like him to bring up. Hence his course was neither a seminar nor a traditional lecture course, but something in-between. The stimulation was many-sided: between the instructor and his students on one side, and students among each other on the other side. In a similar way, I hope by putting these notes of mine on the World Wide Web that readers will be stimulated to react critically either for or against my ideas and will perhaps also exchange ideas with each other.[1]
The issues discussed here are problems to which I shall try as far as possible to abstain from suggesting answers. Many of them are problems to which I know no answers. Perhaps some of them are problems to which there are in the nature of things no completely satisfactory answers. To some extent, crucial problems, in Jakobson's sense, are open-ended and lead to more problems rather than end up being definitively solved. However, as he might have said, they continue to be 'stimulating and fruitful'! I sometimes feel that it is salutary to try as far as we can to rid ourselves of the notion that to every problem there is a unique final solution which it is our duty to find, together with the notion that once we have found these solutions we need to convince our colleagues that they are indeed the only correct ones. There is a certain intellectual arrogance hidden not too far below the surface there! In this connection, I find that linguistic historiography and even linguistic theorizing are too often aimed at producing what seem like incontrovertible statements, purportedly based on empirical results or logical arguments, which it is imagined that readers or listeners will sit back and compliantly accept.
I purposely define linguistics as broadly as possible to subsume (1) the study by native speakers of the languages that they speak, (2) the study of foreign languages or foreign language and dialect varieties, and last but not least (3) research and speculation on the nature of human language and its manifold relations with the rest of reality. Hence, this embraces both theoretical and applied linguistics as those terms have been used for the past century and a half. It also includes all language pedagogy, whether or not the language in question is spoken natively. Under what sort of historical conditions does linguistics, broadly defined in this way, originate? Why do some cultures appear to devote more attention, other cultures less attention, to this vast domain? Is linguistics only carried on in so-called civilized societies? In other words, is there an analogue to linguistics in more traditional societies? How has linguistics in fact originated in well authenticated cases, viz. in western Eurasia, India, and the Far East? Alternatively, should linguistics be defined more narrowly in order to reduce this huge domain and ease the burden on historians of linguistics? After all, the definition of linguistics is itself a prime example of a crucial problem in Jakobson's sense of the term.
If we continue to view linguistics as embracing the entire field of language study as I have just suggested, we can then ask the following questions. Having originated, how has the study of language functioned and operated, and how has it been related to other aspects of intellectual culture? The first question can be amplified thus: What has linguistics involved? What have been the goals of, and what has been achieved by, linguistics in different periods? Specifically, how has it been conducted in practice? What has been the intellectual equipment needed to carry on linguistic studies of these various kinds? The second question can be fleshed out thus: How has linguistics fed into other aspects of intellectual culture, and how has it in turn been affected by it?
To approach the latter question we need to concede at the outset that it is not easy to separate linguistics from overall intellectual culture and treat it as a completely autonomous tradition. This is an especially thorny problem when one considers periods before the advent of any explicit grammatical tradition or of linguistics specifically so named. What is nowadays called philosophy, for instance, has traditionally subsumed a number of linguistic problems (meaning being one of the most notoriously persistent). Can these linguistic strands in the history of philosophy be separated out, and is it profitable and legitimate to do so? How should historians of linguistics themselves handle those aspects of language that have traditionally constituted part of the province of philosophers, not to speak of other specialized fields (rhetoric, the study of speech and communication, etc.)? Recall in this connection that we observe an increasing trend among linguists of the past two centuries to regard linguistics as separate from everything else. How does this trend manifest itself, and what are its origins? How long have linguists regarded linguistics (however named) as an autonomous field, and what sorts of arguments have they advanced to defend such a position? In general, should attitudes of the lay public to language over the centuries be regarded as falling within the province of historians of linguistics? Such attitudes include not only linguistic purism (e.g. notions of correctness), but also all the interpretations and explanations that people are accustomed to assign to linguistic facts (e.g. questions regarding the origins and appropriateness of words). How has language pedagogy been related to more theoretical strands in the field? Should the teaching of basic literacy be considered an integral part of linguistics, and if so how does it relate to other pedagogical activities, such as the teaching of prescriptive grammar, or in more recent periods the teaching of avowedly non-prescriptive grammar?
Is it reasonable to distinguish scientific from non-scientific approaches to language? Can we define 'science' in such a way as to make this question meaningful? Is there, for instance, such a thing as 'the scientific method,' identifiable across disciplines? Some philosophers of science have certainly denied that this is the case.[2] Even if we assume that we know how to distinguish between more scientific and less scientific approaches to language, at least roughly, do these various types of study develop in similar ways? Moreover, how similar in the past has the scientific study of language been to the scientific study of other aspects of reality? Is it possible, if I may play the devil's advocate, that the history of linguistics is more similar to the history of non-scientific activities, such as the fine arts, literature, music, and theology? Should we be wedded to the notion that the two domains of science and non-science are mutually exclusive and jointly exhaustive?
Have the most significant theoretical insights into the nature of language been provided by linguists, i.e. people who happen to call themselves linguists? Especially when we come to the twentieth century, how much stimulation have professional linguists received from other fields? Is it important to investigate the ebb and flow of this cross-fertilization, and if so how should this be done? Has it simply been a question of linguists who have familiarized themselves with the voluminous literature dealing with language produced by philosophers, anthropologists, et al.? In other words, does cross-fertilization take other more personal and more productive forms? To what extent, for instance, have professional linguists actually co-operated with non-linguists on projects of mutual interest and drawn theoretical inspiration in that way? What different forms has this co-operation taken, and what results or insights have been obtained that would not otherwise have been gained?
To what extent has linguistics shown linear continuity of development? Have there been periods of theoretical upsurge? Have such periods sometimes been checked by unfavourable cultural events? To what extent have theoretical insights, once gained, been preserved regardless of surface events (barbarism, cultural decline, etc.)? Alternatively, could the development of linguistics be portrayed as a sine-curve: repeated periods of elevation followed by periods of decline? For instance, what was the fate of linguistics during the Middle Ages? Did the study of language largely grind to a halt after the collapse of ancient civilization, or did it merely go underground to re-appear revivified during the Carolingian period and the High Middle Ages? Similarly, can the Italian Renaissance of the fourteenth and fifteenth centuries be viewed as a sudden fresh elevation, or did it rather build on insights gained during the immediately preceding period? Is the issue of linearity also relevant to later periods of linguistics? Can we distinguish periods of especially creative theorizing or has the study of language always continued at about the same general level of theoretical sophistication?
Should historians of linguistics follow the same methodology as scholars observe in other branches of history? This question raises a number of difficult issues. Here are two: First, to what extent do we observe historians following a definable methodology? Second, what is 'methodology' and to what extent can it be reduced to a system? The latter is definitely a philosophical question, and it is worth pointing out that philosophers are far from agreeing on its proper formulation, still less on a satisfactory solution.
Is it possible to apply the techniques used by historians of science to the history of linguistics? If so, what justification can be offered for extending the techniques of the one to the other? Is it because the techniques of the historian of science are ipso facto more legitimate than the techniques of the general historian? Finally, are such techniques uniquely characterizable, and to what extent can they be imitated? The same question may be posed with regard to the techniques of the general historian. Needless to say, there has been much discussion among historians about the way history is written. See, for example, the influential book by Paul Veyne entitled Comment on écrit l'histoire: Texte intégral, 2nd ed. (Paris: Éditions du Seuil, 1996; an English translation of an earlier edition appeared in 1984). A general historical survey of different styles of historiography may be found in Benedetto Croce, History: Its Theory and Practice, translated by Douglas Ainslie (New York: Russell & Russell, 1960). For a more recent treatment of this problem by an accomplished historian, see Herbert Butterfield, The Origins of History, edited with an introduction by Adam Watson (London: Eyre Methuen, 1981). As regards linguistic historiography, a fine discussion of the basic issues is provided by Peter Schmitter in his Historiographie und Narration: Metahistoriographische Aspekte der Wissenschaftsgeschichtsschreibung der Linguistik (Seoul: Sowadalmedia & Tübingen: Narr, 2003).
How should the results of linguistic historiography be presented? In their books and articles, what should the historian of linguistics focus on? The following possibilities suggest themselves: biographies of individual scholars, exegesis of selected seminal works, reconstruction of disciplinary traditions, careful examination of foundational issues (a topic that the French call épistémologie), reconstruction of actual disciplinary practice, and so forth. Should mixtures of these different approaches be experimented with? What role should philological procedures play in this context? More generally, should philology be considered part of the equipment of the historian of linguistics? Can the historiography of linguistics be conducted without any resort to philology? I am obviously using the term 'philology' here in the sense of textual criticism and analysis, not as a synonym for 'historical linguistics.'
If we define the historiography of linguistics broadly as the study of the history of linguistic studies of all kinds, is it useful to professional linguists? If so, in what way? Are the results achieved by historians of linguistics of greater potential interest to some linguists than others; more so, for instance, to linguists who work on foundational issues than to linguists working on the various sub-fields of linguistics (applied linguistics, grammatical theory, phonology, anthropological linguistics, psycho-linguistics, socio-linguistics, etc.)? Should all students of linguistics be encouraged or even required to take courses in the history of linguistics?
Is the history of linguistics of any value to workers in other fields of knowledge than linguistics? If so, in what way? If not, why not?
In particular, could the history of linguistics be relevant to historians who work in other branches of history, such as political history, intellectual history, the history of science, and so forth? Again, if so, in what way? And if not, why not?
How relevant, if at all, is the study of the history of linguistics to general education? Should undergraduates be required to take courses in the history of linguistics? What place does the history of linguistics have in graduate education?
What is an adequate curriculum for a course on the history of linguistics? When such courses are taught in the New World, for instance, should they deal predominantly with developments on this continent, or should they also attempt to cover the history of linguistics in other parts of the world? Analogous questions may be raised about courses on the history of linguistics presented in other parts of the world. Moreover, to what extent should courses on the history of linguistics focus on theoretical developments since 1900 or on the genesis of theories that are still current?
What is the most valuable secondary literature in the field? What should we advise students to read in order to introduce themselves to the field? How can we steer students towards the best secondary literature in specialized areas? In which languages should a student of the history of linguistics be encouraged to be competent?
How can students be trained to do linguistic historiography? Should a course in the history of linguistics include a component dealing with the general principles of historical research? To what extent is it desirable for students to be trained to handle primary historical sources?
What are the future prospects of the field, and what steps can we take to improve them? What steps should we take to improve international co-operation in the field?
How can members of the profession improve their own historiographical skills? To what extent is faculty development of that kind feasible? Underlying this issue is the question of how well the field is being served at present in Linguistics departments in universities in Europe, Asia, Australasia, and the New World. Can this situation be improved? Can we take steps to ensure that courses on the history of linguistics are offered in more linguistics departments than at present? Can courses on the history of linguistics be organized at summer schools? Crucially, where can we find people willing and able to teach such summer-school courses?
I add an appendix here to show that this kind of open-ended questioning may be applied to a specific problem, namely the issue of the relevance of Thomas Kuhn's much-discussed paradigm theory to the history of linguistics. I might mention here that I myself have discussed this question in print.[3]
First question: Can the history of every scientific discipline be divided into a finite number of discrete phases? This is the quantifiable aspect of Thomas Kuhn's well-publicized notions of paradigm and 'normal science.' Are the boundaries between chronologically adjacent paradigms identifiable?
More fundamentally, are scientific revolutions episodic, or do they instead characterize the entire growth of scientific knowledge? In other words, could it be that revolutions are a constant element in the history of science, as Imre Lakatos was fond of emphasizing?[4]
What is the unit-community within which scientific revolutions take place? Can one delimit and identify such units of analysis? In a field like astronomy, for instance, how many communities are there? Is Kuhn's framework only good at handling situations in which a clearly circumscribed set of theoretical problems is dealt with by a clearly circumscribed group of practitioners, i.e. situations in which a new idea can suddenly revolutionize the perspective of a small group of scientists dealing with a limited set of problems? Do research problems define scientific communities, or conversely do scientific communities define paradigms? Is it not possible that every academic subject embraces a multiplicity of communities and hence a multiplicity of simultaneous paradigms. If so, is that a welcome, or even a tenable, conclusion?
Was Kuhn right in focusing on the 'bandwagon effect'? Do scientists typically rush to embrace each new paradigm as soon as it appears on the scene? How potent has this bandwagon effect been in the past? Is it possible that the bandwagon effect is to some extent an illusion, i.e. a function of our present-day perspective? On the other hand, if the bandwagon effect is indeed genuine, is it possible that it has varied in intensity in recent times? If so, are we nowadays witnessing an increase or a decrease in that effect? How far is the bandwagon effect caused by the way scientific projects are financed? Does government funding of science intensify the effect? To what extent do scientists in fact succeed in avoiding the stampede caused by new paradigms and continue to work independently? In each period, how productive was the work of such non-conformists, and how influential was it subsequently? What has been the role of the counterpart in the scientific community of 'indies'? What has been the role played by serendipity in the history of science? (For a recent study of this phenomenon, see Robert K. Merton & Simona Fallocco, La serendipity nella ricerca sociale e politica: cercare una cosa e trovarne un'altra [Rome: LUISS, 2002].)
Do paradigms in one scientific community affect paradigms in other communities, and if so how is this influence mediated? Does each individual belong to a single scientific community? If scientific communities overlap, what does that entail for Kuhn's theory?[5] Kuhn himself seems to have realized more and more that his scientific communities were inherently vague entities.[6]
What is the relation between the innovating individual (or individuals) and the scientific community that eventually adopts a new paradigm? Does acceptance have to be universal to justify applying the term 'scientific revolution' in Kuhn's sense? What changes if any does a new paradigm undergo in the process of being expounded to the community as a whole and in the process of being embraced by such a community? Are new paradigms accepted unchanged, or do they undergo modification in the process of being disseminated and assimilated? Does the innovating individual have to be a leader in the scientific community in question, and if that is so does that not make the theory of scientific revolutions inherently elitist? If we consider leadership role, should we not also consider amounts of financial support, whether they come from private or public sources?
To what extent is Kuhn's theory applicable outside the natural sciences? Did Kuhn's focus on the natural sciences make that question otiose? I think it is true to say that Kuhn himself was only interested in the sciences in the narrow Anglo-American sense of that term. But perhaps what he discovered can nevertheless be applied outside those disciplines. If that is so, to what extent is that the case? How successful have such attempts to apply Kuhn's theory outside natural science been so far?
What stretches of the early development of the natural sciences did Kuhn cover in Structure? When are we to say that true science starts in the case of celestial mechanics, chemistry, geology, physics, etc.? This raises the issue of 'pre-paradigmatic' disciplines in the sciences. In what respects were they like the present-day humanities? If Kuhn's theory is applicable outside the natural sciences, can one legitimately speak of pre-paradigmatic non-scientific disciplines, or is that not a flagrant contradiction in terms?
What implications do the answers to all these questions have for the development of linguistics and for the way historians of linguistics conduct research on the history of their discipline? To take one example among many, can we identify a pre-paradigmatic stage in the study of language? When did the study of language become a science (if it has ever done so)? If it did indeed become a science, who was responsible for this decisive transformation? Can we identify the innovating individual (or individuals) and the cultural circumstances surrounding this event?
Do linguists today constitute a single scientific community in Kuhn's (admittedly vague) understanding of that term? If not, how many such communities (roughly) can be distinguished in the field of linguistics? Does each individual linguist belong to a single community, or do some (or even most) linguists belong to more than one community? Do linguists in one country belong to the same communities, roughly speaking, as linguists in another country, or are there scientifically distinctive national styles of linguistics? To what extent, therefore, is the linguistics carried on in different countries commensurable? Are some national styles of linguistics more 'scientific' than others (in some reasonable sense of that treacherous term!)?
If we believe that linguistics is now a science, how many paradigms, in Kuhn's sense of the term, has it passed through, and whom do we identify as responsible for each successive revolution? Have revolutions taken place in the history of linguistics for which no unique single innovator can be identified? What has been the role of non-conformist scholars in the development of linguistics?
Is it legitimate to raise the question of the overall direction of change in linguistics over the past two or so centuries? During that period, can linguistics be said to have 'gone in the right direction,' however we understand that expression? Is Kuhn's conception of the development of science as a more or less random, more or less discontinuous process verifiable in the development of linguistics, and if so to what extent can it be reconciled with the notion of progress? Can one identify an ebb and flow in linguistics, either in the distant past or in more recent times, both in the breadth and depth of such studies? Should historians of linguistics feel that they are uniquely qualified to warn linguists of the direction in which the field is moving at the present time, and to what extent is such an attitude justifiable? On the other hand, if they abstain from thinking about these questions or having thought about them remain silent, can they be accused of lacking a proper sense of their scholarly responsibilities?
I welcome comments. Caveat lector! While I have attempted to appear non-committal, the questions that I pose here inevitably conceal presuppositions. Moreover, there may be wider perspectives that I have ignored, and awkward questions that I have failed to raise. For these and other reasons, I welcome all comments, the negative as well as the positive. I thank colleagues who have already responded. Recall, please, that a web document is unlike a publication in book or article form in that it is not fixed and permanent. I frequently revise the documents on this website, so readers might consider re-visiting them from time to time to see what changes have been made.