Bertrand Russell, Language and Linguistic Theory

Keith Green

• Publisher: Continuum
• Year: 2007
• ISBN:
  9781441180339
• Edition: 1
• Topic: Langues et linguistique
• Subtopic: Linguistique

Although there has been a significant revival in interest in Bertrand Russell's work in recent years, most professional philosophers would still argue that Russell was not interested in language. Here, in the first full-length study of Russell's work on language throughout his long career, Keith Green shows that this is in fact not the case.In examining Russell'swork, particularly from 1900 to 1950, Green exposesa repeated emphasis on, and turn to, linguistic considerations. Green considers how 'linguistics' and 'philosophy' were struggling in the twentieth century to define themselves and to create appropriate contemporary disciplines. They had much in common during certain periods, yet seemed to continue in almost total ignorance of one another. This negative relation has been noted in the past by Roy Harris, whose work provides some of the inspiration for the present book. Taking those two aspects, Green's aimhere is to provide the first full-length consideration of Russell's varied work in language, and to read it in the context of developing contemporary (i.e. with Russell's work) linguistic theory.

The main aims of this important newbook, in focusing exclusively on Russell's work on language throughout his career, are to place Russell within the changing contexts of contemporary linguistic thought; to read Russell's language-theories against the grain of his own linguistic practice; to assess the relationship between linguistic and philosophical thought during Russell's career, and to reassess his place in the history of linguistic thought in the twentieth century. As such, this fascinating study will make a vital contribution to Russell studies and to the study of the relationship between philosophy and linguistics.

Chapter 1

Contexts and connexions

‘I like words to cut like a scimitar, clean, deep cuts...’

Letter to Ottoline Morrell, July 1918

‘The distinction between the emotional and the logical use of words is illusory.’

Russell, ‘The Meaning of Meaning’ 1926

It is widely recognized that Bertrand Russell’s writings represent a phenomenal output over more than eighty years. The volume alone is simply staggering: about sixty million words written during his adult life. It is also widely assumed that Russell was deeply suspicious of language’s function in relation to mathematics and philosophy and that he was appalled by language-centred philosophical practice, by philosophy that focused on language rather than the world. The orthodox view of Russell’s philosophy is that it is consistently ‘anti-linguistic’ in character, although this view is based largely on a reading of his relationship with the Oxford ‘ordinary language’ philosophers, influenced substantially by the later Wittgenstein, of the 1950s. Such a reading would also admit that Russell certainly used language, however – and used it prolifically. It is estimated that he wrote an average of two thousand words a day throughout his adult life, and although some of the writing is bound to be trivial and ephemeral, the bulk is of more significant interest. From the earliest letters to his grandmother and Uncle Rollo (circa 1882), to the ‘Greek Exercises’ of his youth (1889-9); from his Cambridge Fellowship dissertation on the foundations of geometry (1892), to his last major work in philosophy, Human Knowledge: Its Scope and Limits (1948); from his prolific outpourings to Lady Ottoline Morrell in the 1910s, to the last open letter protesting about Israel’s air-raids on Egypt just before his death in 1970; from his adolescent essays, ¹ to his last essay on freedom in 1967, Russell’s life was a life in words.

As I have intimated, however, few would suggest that his was a life about words. Yet considerations about language and linguistic theory litter Russell’s works throughout his life. It is often maintained that Russell held no philosophy of language, a view which is once again based on his more vehemently anti-linguistic writing from the latter stages of his philosophical career. But it is clear that he invariably approached philosophical problems with at least an implicit assumption that issues of language were at the heart of the problem, even if he did not agree that such issues were the whole problem. Moreover as Gordon Baker (1988) points out, somewhat startlingly, Russell was pivotal in determining linguistic thought in the first half of the century and beyond. Some linguists may find that judgement surprising, for if we were to list the pioneer linguists of the first half of the twentieth century it is doubtful that Russell’s name would be among them. Rather, we would have Jesperson, Sweet, Gardiner, Sapir, Whorf, Bloomfield and Trubetzkoy, among others. This of course depends on how we read the term ‘linguistics’, but it reveals much about the relation between the two disciplines. One inference to be made here is that Russell was apparently pivotal in a discipline that virtually ignored him and that he himself barely recognized.

A different view altogether is expressed by Ray Monk (1996a), Russell’s biographer, who is convinced that Russell was never much interested in language. If we put the two judgements together we have a view of a philosopher uninterested in language, yet crucially determining the nature of linguistic thought. This situation, if we are to believe it, is brought about in two ways: the mutual ignorance of philosophy and linguistics in the first half of the twentieth century, and the assumption that Russell’s remarks concerning ordinary-language philosophy stand as representative of his philosophy throughout his life. Although Monk’s thesis is not that which is represented in this book, it is easy to see how such a conclusion could be reached, for Russell himself periodically complained of language’s misleading signs. He also, however, very often exclaimed that a focus on language in philosophical issues was essential, and that, for instance, the structure of language can tell us much about the structure of the world. With regard to this issue, however, Russell’s conclusions are far from clear, as I show in Chapter 3; but the fact is that whether he perceived language as misleading or truth-telling, it was something he constantly investigated. And the fact is, of course, that language has always been of central concern to Western philosophy; from the Ancients, in particular Aristotle, who developed subject-predicate logic; to the mediaeval modistae, who looked to grammar as a reliable record of facts; to the Renaissance humanists, who prized language over thought; to the scientific rationalists of the seventeenth and eighteenth centuries, who looked for a ‘rational grammar’ and to the idealists and the logicists of the modernist world. Russell contributes to a long tradition despite the fact that he sometimes, particularly in his most productive period from 1900-10, appears to wish to eradicate linguistic issues from analysis. But the problems of the relation of language both to the world and the philosopher’s vision are naturally in part linguistic: it is almost tautological to say so. What gives rise to the disbelief about Russell’s contribution to the philosophy of language is partly accounted for by the rise of what we now call ‘linguistics’ in the twentieth century, and its separation from philosophy.

Russell certainly produced important work in linguistic theory without ever being committed to the kind of analysis it spawned in the middle and latter half of the twentieth century and without ever acknowledging it as essentially ‘linguistic’. Baker’s judgement of Russell’s language work and its influence, cited above, is very much a philosopher’s judgement, for the linguists of the period in question do not demonstrate their concurrence. Indeed, for much of the early part of the century linguists are as silent about philosophers as philosophers are about linguists. What Baker is really referring to is not ‘linguistic thought’ but ‘philosophico-linguistic thought’, the relationship between the two being one of the main strands of this book. However, Baker has shown one of the many ironies encountered when considering Russell’s linguistic work. The ‘linguistic turn’ taken by Western philosophy during the early part of the century (though some would date the turn after the publication of Wittgenstein’s Tractates), where language itself became the central focus of philosophical study, was in part begun by Russell and then ultimately used against his own theories, both by Russell himself and later contemporary philosophers such as Gilbert Ryle. As Baker further notes, Russell’s best work on language was undertaken when language per se was not explicitly his concern. But linguistic issues are central to his philosophy, whether Russell himself would have recognized this fact or not.

One of the aims of this book is to reassess Russell’s position in the philosophy of language and linguistic philosophy; another is to situate Russell in the context of contemporary linguistic thought. These aims arise from my belief that, contrary to what many critics have assumed, language and linguistic theory are not merely peripheral concerns in Russell’s philosophy. A further, though much less central, aim is to investigate Russell’s own use of language, not merely for stylistic purposes, but in order to show the relationship between linguistic theory and linguistic practice. For those familiar with the so-called ‘Continental’ philosophical tradition, this method of reading a subject’s habitual linguistic practice against his or her considered linguistic theory is well known, but for those working within the tradition of analytic philosophy, it is much less so; and when it is known it is generally disparaged. However, it is not my express purpose to submit Russell’s rhetoric to the scrutiny of Derridean deconstructive practice, or to apply modish techniques of rhetorical analysis to Russell’s linguistic traits. Rather, it is to use certain aspects of deconstructive methodology in order to read Russell ‘against the grain’. Russell offers a seductive binarism in this respect: for Russell, language is Janus-faced in that it can ‘cut like a scimitar’, yet obscure the logical form of things; it can sometimes tell us about the world and sometimes lead us into errors of perception about the world; it is at once the object and tool of analysis.

It is fair to say then, I think, that on the whole Russell was suspicious of language’s role in relation to the phenomenal world, while admitting its crucial importance. Yet, as I have suggested, his was a life in words. I do not mean to imply that Russell was a man who wrote rather than did, however. Time and again in his writings we encounter philosophical programmes such as the refusal to let grammar, ‘dictate to ontology’, and the admission of the centrality of an external world separate from representations of it (‘things are what they are, and there is an end of it’, he says, in ‘Vagueness’, CPBR 9:148). He never allowed a theory of symbolism to deflect him from the attempt to see beyond language; yet the problems of defining an adequate symbolism, of which language is the most important and pervasive, abound in his work. It needs to be remembered that Russell’s most famous work, that ‘paradigm of philosophy’ as Frank Ramsey called it, the Theory of Descriptions, is in part a theory of language focused on a single linguistic element, viz. the definite article. This may be a paradigm of philosophical analysis, but it is also linguistic analysis of a first order too, although as I shall show in subsequent chapters, it is very much a philosopher’s linguistic theory. Although Dummett (1991) locates the ‘linguistic turn’ in philosophy at a precise moment in Frege’s Die Grundlagen der Arithmetik (1893 and 1903), one might also note a completely independent linguistic turn in Russell, albeit a much harder one to pin down, leading up to the publication of ’On denoting’ in 1905. That moment in Frege’s work comes about through a discussion about the nature of number followed by a reflection upon the meanings of sentences containing number words and expressions. Thus the investigation slips seamlessly from mathematics to linguistics, from ‘2’ to ‘the even prime’. The precise nature of this linguistic turn is still being debated, but it seems to foreshadow developments much later in the twentieth century, developments in linguistics and philosophy centring on the meaning of words in relation to their use. In Russell the linguistic turn comes gradually through a number of papers between 1900 and 1905, including the chapter on denoting in The Principles of Mathematics (1903, hereafter PoM), despite the fact that he did not read Frege until he had nearly completed the book. There is a further ‘psycholinguistic turn’ in the 1920s evident in Russell’s writings due in part to the influence of Wittgenstein. Recent revisionist readings of Russell’s place in the history of analytic philosophy, such as Dummett’s (1991), stress his antipathy to ‘linguistic’ analysis. Similarly, Nicholas Griffin (1996) and Harold Noonan (1996) both see the Theory of Descriptions (in ‘On denoting’) as a tool for precisely eliminating linguistic analysis and maintaining the analytical focus on the world.

Although this is in some respects a plausible claim, I suggest that Russell’s involvement in, and need for, linguistic analysis is too evident to be so readily discarded, and the analysis of the principles of denoting in Chapter 2 are thus a modification of this revisionist thesis. Monk (1996a, 1996b) considers that after 1919 Russell adopted a ‘linguistic’ view of analysis which transformed his philosophical work; but this is not the true Russellian linguistic turn. Rather, it is a later development reluctantly undertaken in response to the work of his onetime pupil, Ludwig Wittgenstein, and one to which he was never fully committed. Here, after 1919, we have a variation on Russell’s linguistic turn, one which can in some ways be seen as a development of his earlier work. What is different is his grudging acceptance of the linguistic character of logic and his move towards a psychological model of propositions. Russell’s need for linguistic theory is quite evident from the early works where he attempts to construct a perspicuous language for the description of mathematical processes. This is not the whole story, but it is where the linguistic impulse begins, and explains why in PoM there are a number of chapters on linguistic issues, including one on ‘Proper names, adjectives and verbs’ (42-52). Not only did he clearly work in explicitly linguistic areas, but he admitted, albeit at sporadic phases of his work, that, as I have suggested, ‘partly by means of the study of syntax, we can arrive at considerable knowledge concerning the structure of the world’ (1995 [1940]: 347). Similarly, in the earlier Outline of Philosophy, he states that ‘the subject of language is one which has not been studied with sufficient care in traditional philosophy’ (1993 [1927]: 35). It can be inferred that Russell is referring to the logical structure of language here, but of all the statements about the workings of ordinary language made by Strawson (1950), the philosopher who mounted the first serious attack on the Theory of Descriptions, the only one he could accede to was that it ‘had no exact logic’. The logic to which he refers is that which is masked by language, but that does not imply that we cannot, through analysis, see behind the mask into the structure of the world. The question is whether the vision is to be had by virtue of a theory of symbolism, which may or may not entail psychological models, or a theory of logical form. As Russell discovered, it is often difficult to separate them.

Although he is often accused of a yearning for a logically perfect language, supposedly showing his misunderstanding of the functions of language in its everyday role, Russell’s conception of linguistic communication is thoroughly empirical. Even in his early work, contrary to what some commentators have suggested, Russell was aware of the functions and structure of ordinary language (that is, natural language in its regular communicative forms) and alert to the issue of linguistic usage. In PoM he talks of different uses of words having the same logical form and of the fixed way in which words are employed (to generate meaning). In An Inquiry into Meaning and Truth he states:

Some modern philosophers hold that we know much about language, but nothing about anything else. This view forgets that language is an empirical phenomenon like another, [sic] and that a man who is metaphysically agnostic must deny that he knows when he uses a word. (1995 [1940]: 347

Language is thus not to be divorced from metaphysical and ontological considerations. What separates Russell from philosophers who generally followed a later Wittgensteinian theory (or anti-theory) of language is his increasing acknowledgement, often reluctant, of the role of psychology and epistemology in particular, in his conceptions of logic and, in parallel, in language. These developments in Russell’s thinking were, as Elizabeth Eames (1989) has suggested, largely rejected by post-war philosophers, who no longer felt the need to attempt to synthesize language, psychology and logic, but gave themselves over wholly to the analysis of only one of the triad: language.

Russell, philosophy, linguistics

Despite the existence of an enormous amount of linguistic work undertaken by Russell, whether obliquely or directly, it is easy to consider Russell, as I have suggested, as ‘anti-linguistic’. This phrase is misleading, however. Although it suggests a philosopher opposed to linguistic considerations, it should be read as referring to someone who is opposed to the kind of linguistic inquiry that was dominating philosophy in the post-war era. Ray Monk, for instance, in a discussion of the foundations of analytical philosophy, states:

Throughout all the various transformations of Russell’s philosophical doctrines, one thing remained quite constant, and that was the conviction that whatever it is the philosopher is concerned with, it is precisely not language. (Monk, 1996a: 4)

It must also be noted that Russell’s most ‘linguistic’ work, An Inquiry into Meaning and Truth (1940), however, has had relatively little influence. This is due in part to the contemporary influence of Wittgenstein and to the rise of descriptive linguistics in its quasi-scientific Bloomfieldian form during the 1930s and 40s. However, despite its often rather headstrong psychologism, it is the clearest mature expression of Russell’s theories of language that we have. Wittgenstein’s near-contemporary work, little of which Russell would have read at the time (although he was sent the manuscript of The Blue Book in the autumn of 1935), culminating in the Philosophical Investigations (1953), did not impress him. For Russell language was not a game or variety of games but an unreliable tool which could, given the correct analysis, tell us something obliquely about the world. Ultimately, in his despairing ‘retreat from Pythagoras’ Russell felt that mathematics itself was linguistic. Yet even these somewhat opposing views – that language can or cannot tell us anything about the world — suggest a Russell troubled by a central linguistic issue. On the contrary, Russell is notoriously difficult to sight because he is such a moving target, a thinker who changed his mind on a number of key philosophical issues and was not afraid of doing so. His shifts of opinion vary in force and importance as well as in topic and content, from his ‘sudden conversion’ to a form of pacifism and anti-imperialism during the Boer War, to his gradual move away from the work and personality of Ralph Schoenman in the 1960s; from his rejection of Hegelian idealism (initiated by G. E. Moore) to his abandonment of his theory of knowledge in 1913 (following Wittgenstein’s severe criticisms); from his acceptance of Wittgensteinian atomism to his return to empiricism in the 1940s (specifically in An Inquiry into Meaning and Truth). Similarly, as Eames states, Russell ‘always looked at his work as tentative and provisional and did not draw back from publishing work in which he knew there were unsolved problems’ (1989:159).

There are some aspects of his philosophy, however, that changed very little, some of these being the more ‘linguistic’ in focus. Russell defended his theory of descriptions against the attack by Strawson (1950) after nearly half a century of general acceptance by the philosophical community. I refer only to the philosophical (rather than linguistic) community here because Russell, in common with other analytic philosophers of the first half of the twentieth century, and as I shall show in greater detail in subsequent chapters, took very little cognizance of contemporary linguistic theory as expounded by linguists and philologists themselves. Perhaps the most influential of linguists, Saussure, is never referred to by Russell. Similarly, although some thinkers associated with both Russell and Wittgenstein, such as C. K. Ogden, were familiar with the Cours de linguistique générale (1917), it appears that linguistic and philosophical considerations of language were strikingly independent of each other.² Russell’s lack of acknowledgement of Saussure can at least be in part attributed to Ogden and Richards’s rejection of Saussurean linguistics in The Meaning of Meaning (1923). In a similar manner, the authors dismiss Michel Bréal’s semantic theory (although admittedly, this is a semantics of a very particular kind), developing in the early part of the century, as being one of many ‘interesting though subordinate fields of investigation’ (Ogden and Richards, 1972 [1923]: vi). But, to stress again, this does not mean that language was not of fundamental importance, and the story of the relation between linguistic and philosophical theory is nevertheless a fascinating one. Philosophical discourse had always assumed that part of its armoury and metalanguage were linguistic terms, and it was happy to use terms from philology and grammatical theory. Philology itself, however – burgeoning linguistic theory influenced by nineteenth-century scientism – was in one sense trying to free itself from philosophical discourse and develop itself as an independent discipline. As late as 1932 Alan Gardiner, in The Theory of Speech and Language suggests that linguistics was struggling to be free of its philosophical burden:

Much more questionable is, indeed, the claim of the philologist to construct a linguistic theory without the help of experts in those abstract fields, [that is, philosophy]. (Gardiner, 1932: 7)

Interestingly it is philosophers who are seen as having ‘linguistic theory’, while philologists must founder without their help. Gardiner further rather ironically suggests that ‘every intelligent workman in any of these branches [philosophy] should possess a shrewd idea how the mechanism with which he is particularly concerned achieves its ends’ (8) – recalling Russell’s acute comment that Wittgenstein should at least have a smattering of logic if he is to provide a foundational theory for it. Gardiner suggests that in the middle part of the twentieth century, philosophers no longer had the knowledge appropriate for ce...

Table of contents

• Cover
• Title Page
• Copyright
• Dedication
• Contents
• Preface
• Chapter 1: Contexts and connexions
• Chapter 2: Language and the -world
• Chapter 3: Language and the mind
• Chapter 4: Language and linguistic theory
• Chapter 5: ‘These fragments I have shored against my ruins’: Russell and Modernism
• Notes
• Bibliography
• Index