The Logic of the History of Ideas

Mark Bevir
Cambridge: Cambridge University Press. (1999)
ISBN 0 521 64034 2 (Hbk.). £37.50, $59.95

A round table book review with contributions from:

1. Alun Munslow
   Professor of History and Historical Theory, University of Staffordshire;
   UK Editor, Rethinking History: The Journal of Theory and Practice
2. Siep Stuurman
3. Allan Megill
   University of Virginia
4. Kari Palonen
5. Frank Ankersmit
   Professor for Intellectual History and Historical Theory
   Groningen University (The Netherlands)

NB This exhange originally appeared in Rethinking History: The Journal of Theory and Practice, 4:3 (2000) (http://www.tandf.co.uk/journals/routledge/13642529.html)

1. Alun Munslow

Mark Bevir establishes the framework and rationale for his tremendously ambitious book on the first page of the Preface (p. ix) when he says he draws on analytic philosophy to study the forms of reasoning appropriate to the history of ideas. He privileges what he calls logical and normative (as opposed to historical, sociological or psychological) analysis later referring to it as a post-analytic philosophy or anti-foundationalism. This procedure is given priority over the ontological hermeneutic tradition (chiefly of Gadamer) and provides his chosen route to the justification and explanation of understanding. He is careful to disclaim that what he is doing is incompatible with that other tradition arguing they can be complementary as different approaches to different issues. In contrast to the Skinner-Pocock school of textual interpretation Bevir argues no single method can logically prise open meaning objectively. Based on this belief is his reading of epistemologists and philosophers of the mind like the later Wittgenstein and Davidson.

In pursuing his substantial objective - to explain the logic of the history of ideas - in Wittgensteinian fashion the author attempts first to establish the reasoning and concepts associated with it. Bevir calls this the grammar of its concepts that can be determined by both deductive and inductive arguments (p. 2). Very quickly Bevir lays his cards on the table with his claim (in the context of his explication of the failings of analytical philosophy) that 'all our knowledge arisesin the context of our particular web of beliefs' (p. 5). Following Rorty, Bevir suggests that especially in its logical-positivist incarnation, all roads lead away from analytical philosophy (and Descartes and Kant), towards anti-foundationalism. As he says, he will 'go along with the anti-foundational (or post-analytic) conclusion that there are no given truths' (p. 6). However, while he eschews 'the given' both empirically and rationally, or any ultimate or privileged representation, he draws back from and positively rejects 'the irrationalist anti-foundationalism found in post-structuralists and post-modernists such as Jacques Derrida, Michel Foucault, and Jean François-Lyotard' (p. 6). This is the essential pivot of Bevir's position. In his effort to demonstrate through his version of the logic of the history of ideas, in which he blurs the distinction between synthetic and analytic propositional forms of knowledge, he attempts to construct and walk a middle road.

This middle road is predicated on Bevir's insistence that the logic of a discipline consists in a normative account of its logic or reasoning, and that invoking historical examples of reasoning simply serve to confuse the issue. In his analysis (of the abstract nature of logics) Bevir claims books like Collingwood's The Idea of History and White's Metahistory confuse the matter by ignoring the logic of explanation which is, of course, the logic of justified belief (deduction and induction). What Bevir tries to do, therefore, is provide a single form of justification that will compromise the synthetic/scientific/empirical and analytical/philosophic/rational forms of knowledge. Because this cannot be done by appeal to 'the given' or accurate representation, it can only be done by an appeal to 'the nature of our being in the world' (p.18). This is explained by a rebuttal of Hayden White's position that historians have no rational grounds for choosing one philosophy of history over another. Bevir suggests White's judgement that our reasons are aesthetic/tropic rather than, as White has recently suggested, logico-deductive (White 2000: 393) should lead him to accept that his choice of explanation in fact commits himself to a particular logic (which undermines his scepticism about knowing truthful things).

The strategy Bevir deploys to pursue his aim, which he does with a doggedness that is remarkable, a clarity of thought which is to be much admired, and a belief in his own abilities that is often breathtaking, is to offer chapter length examinations of the concepts of meaning, objectivity, and belief (the objects of study of the historian of ideas and the essential elements of his grammar of concepts), and synchronic and diachronic forms of explanation and what he calls distortions. By these examinations, and through the application of his 'logic', Bevir offers a valuable insight into the nature of historical thinking and its rethinking.

In his first exploration called 'on meaning', he addresses the nature of intentionalism (strong and weak versions) concluding that he has provided the 'core of a theory of historical meaning' that derives from an individual's weak intentionality in their individual utterances/viewpoints (pp. 76-7). How can the historian of ideas reconstruct the weak intentions of agents objectively? How can we really know what the intentions of the utterer were? By opposing objectivism and scepticism (fixed as the extremes of post-modernist relativism and irrationalism and foundationalist objectivism) Bevir concludes there is a middle road based on 'human practice'. This is Bevir's recognition that while we cannot be certain as to the grounds for what is good history, we can be reasonably sure about our knowledge of past intentions his definition of objectivity by comparing and contrasting competing rival webs of theories about meaning. We can also relate objectivity to truth because of what he calls his anthropological epistemology, that is, as he say 'because our ability to find our way around the world vouches for the broad content of our perceptions' (p. 109). Again Bevir walks but more assuredly now - the middle road. Following on from his position that just because historians are implicated in the act of historical knowing, it does not mean they cannot have objective knowledge of the past, Bevir moves to his analysis of belief. Working from his argument that we can justify objective knowing through fallible 'human practice' and his intentional theory of historical meaning Bevir concludes that sincere, conscious and rational beliefs can be entertained if one does not also hold absolutist expectations of truthfulness. Hence we can explain intentionality in terms of the reasonable expectation of individuals holding such beliefs.

Bevir next explains the nature of the forms of explanation deployed within the history of ideas first what he calls synchronic explanation. This he argues is the formulation or description of the webs of belief held by historians of ideas. This he explains with reference to the connections between tradition and agency. Bevir then attempts to explain how people develop, depart from and change their (inherited, traditional) webs of belief historically, i.e., diachronically. Because people have agency the historian of ideas must have an explanation that accounts for the exercise of that faculty. Bevir uses the concept of the dilemma to explain how rational people change their minds and adapt/adopt new beliefs/webs of belief while remaining sincere, conscious and rational. Bevir concludes his study with an assessment of the irrational, unconscious and deceitful distortions of belief that intrude upon any explanation in the history of ideas. He suggests historians can explain such distortions (deception, self-deception and irrationality) by first recognising such distortions and arguing they arose as logical consequences of the grammar of his concepts. Defined as rogue pro-attitudes explained within the folk-psychology of reason, desire or need. People will the distortions.

Bevir concludes his grand tour of the grammar of concepts and forms of explanation that deploy them by applying his logic to his own explanation. This is Socratic undertaking and an interesting procedure, and one I applaud up to a point. In a sense of course it is just a rationalisation within the terms established within the book and it cannot, therefore, be taken a serious reconsideration, and it certainly isn't a refutation of itself. The implications for the broader field of historical studies of Bevir's explanation of distorted beliefs cast as it is within a framework of rational action theory, is interesting in that it is, as I am sure many historians will point out, a statement of the blindingly obvious (p. 316). Historians 'know' from their own experience of folk psychology how deviancy can be explained and used to explain people's actions at a 'common sense' psychological level. Do historians need to be told how people behave in the ways described and codified by Bevir?

In his rejection of the given empirical or rational Bevir pursues his own middle of the road logic or grammar of those concepts operating in the discipline against a background of webs of beliefs. This activity of the justification of meaning and explanation is neither material nor linguistic, it is conceptual. The obvious question is whether you can have a logic or grammar of the concepts used in a discipline, especially history. The other question is, assuming you can have a grammar, has Bevir described it? Are there others? Indeed it is possible to offer detailed criticisms of each of the concepts as defined by Bevir in his grammar. Does he, for example, give adequate attention to Nietzsche's critique of objectivity, i.e., perspectivism? Why should we accept the basic premise of rational action theory? Bevir would, of course, say not to means falling into the miasma of post-structuralism and that is irrational.

Many historians will share Bevir's anti-foundational, liberal and non-reductionist position. Almost certainly the majority of historians would agree with him that our existence is affected/mediated by our concepts and that there is no absolute extra-discursive ground for knowing, that all texts are interpreted within the skein of our webs of ontological commitments, and they would certainly endorse his dismissive attitude toward the 'irrationalism' of postmodern approaches. What would make them less happy is his rejection of the correspondence theory of knowledge. These historians what elsewhere I have called constructionist historians would be unhappy with his assumption of rational action theory as the centre of historical understanding. Placing the individual making rational choices in a chaotic world at the centre of historical explanation is for most of them, frankly, unconvincing. To then build a whole logic of historical explanation on it is juvenile. To be more charitable, however, Bevir's position is simply very unfashionable. At worst it denies the role of structure, power, and the embedded nature of irrationalism in both motivation and argument (not as a descent from the ideal but the everyday practice of historical agents) in understanding and explaining the past. I suspect the majority of historians would say Bevir's efforts, though clever, could only be written by a non-practitioner or, even more alarmingly, by the naïve reconstructionist members of the profession.

Central to Bevir's undertaking is the very important question of agent intentionality. He seems to be defending the idea (hermeneuticist in inspiration) that we can accurately interpret the author's meaning when they wrote a text. He roundly attacks Pocock's and Skinner's view that we should take into account context and language rather than authorial intention in doing this (semantic as opposed to hermeneutic meaning). Bevir comes down on the side of 'weak intentionalism' that is the middle of the road position. So it is he establishes his ramparts against all-comers from post-structuralists to materialists. Bevir is not afraid of a fight. This is a good thing as the flaw in his argument (which can be pointed out by all and sundry and no doubt will be) may be that his equation of meaning with a weak (or strong for that matter) version of intention. In a nutshell, it is possible to argue that just because an utterance is made, it need not necessarily express an intention. Knowing the intention of the author is of no use in determining what the text means unless you accept that intention does equate with meaning. Arguably, in history we have to make up our meanings unless we believe they pre-exist in the data/text and we can, therefore, 'discover' them.

For all its complexities and what for many will be his failings, not least Bevir's unqualified belief in rationalism over irrationalism, sincerity of insincerity and the conscious over the unconscious, the questions his book addresses are important to historians in their everyday work. Of course, those historians who harbour an anti-theory bias will never read it. The very title will put them off. Locate 'logic', 'history' and 'ideas' together in a book title and it is the kiss of death so far as most jobbing historians are concerned. For the majority, I fear, being a good historian does not mean knowing anything about logic, much less the history of ideas. But such a pre-judgement would be quite wrong. Bevir's explorations are useful reminders of the complex nature of agent intentionality, objectivity and belief. From the middle of the road Bevir is able to offer assistance to those who want to believe in the (more or less) accurate knowability of the past. Of course, if you walk down the middle of the road you are also very likely to get knocked over. But I think Bevir knows that and is ready to take the risk.

References:

White, Hayden (2000), 'An Old Question Raised Again: Is Historiography Art or Science?', Rethinking History: The Journal of Theory and Practice, 4.3, 391-406.

Munslow, Alun (1997) Deconstructing History, Routledge: London and New York.

Discourse index | back to the top

2. Siep Stuurman - On Intellectual Innovation and the Methodology of the History of Ideas

I. Introductory observations

Mark Bevir's Logic of the History of Ideas is a highly ambitious and thought-provoking book that forces all of us to rethink the methodological rules we routinely employ in our research. In this comment, I shall approach Bevir's 'logic' from the vantage point, not of a philosopher, but of a practising historian of ideas. In my opinion the history of ideas is a craft rather than an enterprise in applied philosophy. I say this in full awareness of the philosophical dimension necessarily involved in all history, and perhaps in the history of ideas in particular. More specifically, I share the anti-foundationalist position adopted by Bevir: our experiences are always mediated by our concepts, their is no neutral ground independent from discourse (even though there is a reasonable limit to scepticism about 'facts'). I also share his Wittgensteinian view of language: texts are only intelligible in the context of a Lebenswelt. Finally, I largely agree with his reservations about post-modernist and neo-Nietzschean attempts to reduce the history of ideas to the interplay of power and rhetoric.

I cannot, however, entirely agree with the rationalist and individualist bias of Bevir's theory of the formation of ideas. Bevir repeats over and over again that the historian must explain the formation of 'sincere, conscious and rational beliefs'. Now I submit that these are extremely important, perhaps pivotal, aspects of belief, but that they do not exhaust the historical record: Not all sincere and conscious beliefs have to be rational (certain religious and mystical beliefs are a case in point), not all sincere beliefs need to be fully conscious, let alone rational, and finally beliefs can be rationalized as well as rational (a distinction that is of vital importance in the analysis of the interaction of language and power).

II. Rationalistic and individualist elements in Bevir's logic

Bevir's logic is rationalistic in at least four senses:

1. Bevir pays scant attention to the role of power. However, ideas are always articulated in a virtual field that is, among other things, structured by relations of power and dominance. Moreover, ideas themselves represent modes of power. Even at the basic level of terminology, power impinges on the use and phrasing of the elementary concepts of discourse, as well as the avoidance or softening of certain concepts and ideas.
2. Bevir constantly emphasizes the pivotal importance of the logic of ideas, which turns out to mean chiefly their consistency or coherence that has to be uncovered and explained by the historian. However, he seems much less concerned with the factors that make for inconsistency and incoherence. To approach a text from the vantage point of consistency is a choice made by the historian, but the same historian may legitimately look for inconsistencies, contradictions and evasive language which equally call for an explanation.
3. The privileged position of consistency in Bevir's logic is based on another rationalist assumption, the thesis that people are conscious authors of the texts they write. As a rule, however, people relate consciously and rationally to part of their discourse, not to all of it: a part of their conceptual vocabulary is taken for granted and not consciously reflected upon.
4. Finally, Bevir's approach is rationalist in the sense that he tends to identify the writing of a text with the construction of an argument. By contrast, he neglects the rhetorical aspects of discourse. In his logic, there is a lot about argumentation and consistency, but nothing about hyperbole, irony, tongue-in-cheek, sleight-of-hand, mock innocence, veiled threats, guilt by association, and the like. In this connection, I would rather adopt Quentin Skinner's than Bevir's position: Skinner's notion of the 'point' an author is making enables the historian to examine how an author shifts back and forth between a general and a specific level of discourse in order to establish a 'theory' and simultaneously to make a particular 'point', as well as to inquire into the rhetorical techniques deployed in such a discursive move.

Furthermore, Bevir's logic of the history of ideas is individualist in at least three senses:

1. Texts are usually written down by individual persons, so much I readily grant. However, the virtual power field mentioned above is present from the very beginning. There is always an intended or imagined audience. There is, so to speak, an invisible censor sitting on the author's shoulder (During the greater part of history there also was, of course, a visible censor residing somewhere else). In short, Bevir's logic tends to underplay the presence of the social within the discursive practice of the individual (note that I'm not saying 'the social character of the individual'!).
2. Many of those who produce texts are engaged in some sort of ongoing conversation with a number of, let us say, 'intellectually significant others'. Many texts are, therefore, far more of a collective product than their factual authorship would have us believe. In Bevir's discussion of the history of ideas, there is no conceptualization of this aspect of authorial agency, nor of the role of intellectual institutions and the ways in which they reproduce or transform ideas in the longue durée of intellectual history.
3. Languages, vocabularies, terminologies persist over time, sometimes over a very long stretch of time indeed. Languages are deployed (and modified, I'm not denying that) by individuals, but the nature of language is social. In other words, languages partake of what Anthony Giddens has christened 'the duality of structure' (Giddens 1979). Authors make arguments or 'discursive moves', but they do so within a pre-structured lexical and semantic field which limits the range of possible utterances: at any moment in time, this field can be theorized as both 'given' and 'constructed'. In contrast to Bevir, I therefore highly value John Pocock's emphasis on the structuring power of political (and other) languages, as well as the various approaches in conceptual history ('Begriffsgeschichte'), because they enable the historian to reconstruct the dialectic between authorial agency and 'given' discursive formations.

In the rest of this comment I will illustrate the above objections to Bevir's logic, focussing on the problematic of intellectual innovation.

III. Intellectual innovation and conceptual change

In a broad sense, all history of ideas is about intellectual change, about the ways in which people rework, refashion and readjust their notions of the world and their reflexions on their notions of the world (their 'meta-notions' of the world, one might say). This process of readjustment goes on all the time, even though most of the time most people are not consciously engaged in intellectual change. Insofar as people write down their ideas more or less systematically, there is a 'logic' to it: people often seek to refashion their thoughts to give them more coherence, at least for themselves. Frequently there is also an instrumental rationale at work: people may, quite rationally, modify the formulation of their ideas in a certain manner because they pursue definite goals, or because they are advocating or combatting a cause. Besides, their discourse may be structured by emotions: people often rephrase their ideas in definite ways because they are moved by love, animosity or fear. Finally, some people sometimes attempt an overall reassessment of their ideas, and this may result in the kind of major conceptual overhaul the historian calls an 'intellectual innovation'.

Bevir explains intellectual change as follows: Historians must 'make sense of the way people modify their web of beliefs by portraying the new beliefs as responses to dilemmas confronting the old ones' (Bevir 1999: p. 263). Elsewhere, he posits that 'a dilemma is a new belief, where any new belief, merely by virtue of being adopted, poses a question of the web of beliefs into which it is inserted'. He then explains that 'a dilemma arises for individuals ... whenever their reflections in relation to their experiences lead them to adopt a new understanding as authoritative'. Next, he insists (correctly) on the theoretically mediated nature of all experience, but adds: 'although, as we have seen, the theorising involved need not be especially conscious or prolonged' (Bevir 1999: pp. 243-4). Finally, he posits that in the making of a dilemma there is a reciprocal non-deterministic influence between social life (experience) and beliefs (entering into experience). Let me summarize: people live through novel or unsettling experiences, they make sense of these experiences by means of new ideas which then have to be incorporated in their existing web of ideas, thus creating a dilemma.

In the rest of this comment, I shall formulate a number of criticisms of Bevir's account of intellectual innovation.

1. I submit that Bevir's explanation of intellectual change in terms of dilemmas is fundamentally ambiguous because the thesis that all experience is theory-laden admits of two interpretations: the 'theory' in question may be, 1. conscious and rational, or 2. tacit and a-rational. The first alternative fits into Bevir's overall logic of the history of ideas, the second, however, does not. From Bevir's own admission that the theorising involved 'need not be especially conscious or prolonged', it follows that the second, more indeterminate alternative is as likely as the first one. I conclude that tacit ('pre-conscious' or half conscious) and a-rational elements can, and often do, play a role in the formation of new beliefs. Once the new belief is formulated, the person in question can of course attempt a rational overhaul of his or her entire web of ideas, but the process of intellectual change cannot be entirely theorized in terms of such a rational reconstruction ex post facto.

   Moreover, we should not too easily assume that the rational overhaul will be successful, or that it will even be fully rational. It is also possible that the new idea will be welded to the existing web in a largely ad-hoc manner. In this connection, it is worth noticing that Bevir's image of the 'web of ideas' is itself ambivalent. If we take the web to be a coherent network of ideas, it fills the rationalist bill; but if we picture it rather as a quilted patchwork, it need not do so. And in many cases it may be truly hard to distinguish between rational overhaul and bricolage.

2. My second criticism concerns Bevir's lack of distinction between explanatory notions and moral ideals within the discourses the historian investigates. The image of 'new experiences / new ideas' works quite well in the first case: an author encounters new phenomena or events and has to adjust his explanation of 'how things are' accordingly. However, the case of the formation of moral ideals is more complicated. Moral notions frequently acquire their poignancy in the course of unpleasant or cruel experiences, they are not so much a reflection upon those experiences as a psychological process of coming to terms with them. For example, authors who have suffered persecution often develop a virulent hatred of oppression and a corresponding love of freedom. The idea of freedom itself, is, of course, not invented by them, but the fact that they feel so strongly about it, their 'moral semantics of freedom', can only be explained by the way they have coped with the personal experience of oppression.

   More generally, the discourse of a particular author is suffused by experientially grounded 'structures of feeling' (a term I borrow from Raymond Williams). Consider the following lines spoken by Macbeth after the evil deed is done:

   ... better be with the dead
   Whom we, to gain our peace, have sent to peace,
   Than on the torture of the mind to lie
   In restless ecstacy ...
   (William Shakespeare, Macbeth,
   III.2, lines 19-22).

   These lines represent a reflexion upon the murder Macbeth has committed, but the express far more than just a rational evaluation of what has happened: above all, they convey the inner torment of an anguished and restless mind (in Skinner's terms, that is the 'point' Shakespeare seeks to make in these lines).

   To cite Macbeth in a discussion of intellectual innovation may seem bizarre, but it is not so. The lines quoted above signify the emergence of a new consciousness of his own identity on the part of Macbeth, to be precise, a consciousness of his own identity as a murderer. They certainly express rational ideas about the nature of a guilty life, bereft of hope and gaiety, but their effective meaning is only recoverable in terms of the structure of feeling they convey.

   The relevance for an analysis of intellectual innovation is then clear, for new ideas often emerge in the context of anxiety and emotional turmoil. Especially those authors who formulate really novel ideas are frequently afraid of their own intellectual audacity, they feel compelled to question their very identity and what they stand for ('am I still a Christian when I say that God and nature ...?'). Their moral sense of personhood is at stake, as if they are standing on the brink of an abyss, and the true, deep meaning of their words cannot be fathomed without taking into account their anguished state of mind. In such cases, the emotional subtext enters into the meaning of the text. Bevir tends to equate psychologistic explanations with 'the ways in which the unconscious and irrational workings of the mind can prompt people to form new beliefs' (Bevir 1999: p. 237). He conceives of psychology in terms of 'dark forces within'. What I am arguing, however, is that emotions, conscious or not, are present within the new ideas they helped usher in: They are part of the text, not just of its antecedent circumstances.

3. My third criticism concerns the effects of power and fear, also in those cases involving only purely rational ideas. Bevir never pauses to realize that most authors during the greater part of human history have written under regimes of censorship (political, religious, mostly both). Furthermore, authors often feel that they have to thread softly and carefully even in liberal-democratic regimes where no censorship exists but many kinds of subtle sanctions are employed against those who 'cross the line'. To uncover the traces of authorial prudence in a text, the historian must engage in an extremely careful, contextual close reading, looking at the interconnections between content, terminology, and rhetoric. In many cases, a contextual reading is also necessary to gauge the meaning of definite silences in a text, that is, to recover the particular rhetoric of avoidance employed by the author.

   Likewise, authors may use willful ambiguity in order to produce a meaning that can be understood (by the 'sympathetic' reader) but that cannot be attributed to them in a straightforward way (by the 'ill-intentioned' reader, or the censor). Bevir has an entire chapter on 'distortions' of discourse, about 'how historians should explain insincere, unsconscious, and irrational beliefs' (Bevir 1999: p. 265). However, that is not the question I'm raising. I submit that a discourse may be willfully 'distorted' for perfectly sincere, conscious and rational reasons. Skinner's book on Hobbes' rhetorical strategy is a perfect case in point (Skinner 1996). Hobbes' use of definite ironical tropes was meant to convey a mocking, dismissive attitude to religion to the 'knowing' reader. All the while, Hobbes affirmed his belief in God in quite direct language. I would call that a rational strategy of 'distortion'. Hobbes pursued two goals at the same time: first, to share with his intended, 'good' readers an intellectual contempt for all revealed religion; and second, to ensnare his potential censors in a maze of words so that they would never be able to prove that he was an atheist, or, for that matter, to decide that he was not an atheist (even today, historians cannot agree on the question whether Hobbes 'really' believed in God or not). The relevance of the case of Hobbes for a discussion of intellectual innovation is, of course, beyond dispute. Generally, the Enlightenment is replete with authors who use this sort of ironical inversions and 'distortions' in their literary strategies.

4. My fourth criticism concerns a deceptively simple question: how do new ideas actually arise, where do they come from? Perhaps it is more a question than a criticism, for Bevir's argument is not entirely clear to me. I his chapter on synchronic explanations, Bevir seeks to refute contextual explanations of intellectual innovation, using Jevons' formulation of the theory of marginal utility as his example. 'Contextualists', Bevir says, 'define the relevant context as a social paradigm, so presumably they will scurry off to study works by Jevons' predecessors, such as J. S. Mill, or people who influenced him, such as Jeremy Bentham'. This approach is bound to fail, he continues, for 'Jevons did not conceive of economic value in the same way as did either the classical economists or the utilitarians'. What historians should do, Bevir then explains, is to investigate Jevons' web of beliefs and find out what he thought about economic value; they will then discover that he developed a marginal utility theory of value (Bevir 1999: p. 217).

   Going by what I know about the practice of the history of ideas, Bevir's contextualists are typical straw men. No contextualist would be so naive as the imaginary historian portrayed by Bevir. In the first place, the historian of ideas will not approach utlitarianism as a single unified discourse, but as a set of discourses. Now Jevons was no utilitarian, but he surely took the notion of an individualist calculus of pleasure and pain (utility and disutility in later economic parlance) from Bentham (Roll 1966: p. 379). Furthermore, the way in which Jevons formulated the problem he wanted to solve was shaped by his critique of what he himself termed 'the Ricardo-Mill economics': the edifice of classical economics was already crumbling, especially since John Stuart Mill had accepted the essential correctness of Thornton's critique of the wage-fund theory, just two years before the publication of Jevons' Theory of Political Economy. Apart from that, Jevons was aware of the greater importance attached to consumer demand and subjective utility in continental economics (Hutchison 1972: pp. 450-53, 466-7; Winch 1972). The contextualist historian is thus led to conclude that Jevons took a considerable part of his conceptual apparatus from previous economic theories as well as from Bentham's felicific calculus. The idea of marginal utility itself was a novelty, even though more 'primitive' versions of it can be found from the mid-eighteenth-century onwards (Schumpeter 1957: pp. 181-2). Jevons was aware of some of these earlier views, notably of Condillac's anticipation of marginal utility which he called 'original and profound' (Schumpeter 1954: pp. 175-6). However, the contextualist historian's job is not yet finished: knowing Jevons' fundamental conviction that economics should become a quantitative science based on mathematically formulated theories, and also knowing that Jevons got his early academic education in natural science, our contextualist will naturally observe that marginalism is basically a form of differential calculus (a mathematical idea with applications in physics pioneered by Newton and Leibniz as far back as the seventeenth century).

   In this way, the real contextualist historian, as distinguished from Bevir's straw man with the same name, might go far towards an explanation of the formation of Jevons' economic theory without invoking the deus ex machina of a totally new idea. There is a more general point involved. I would submit that intellectual innovation is frequently, perhaps mostly, not about totally new ideas, but about new combinations of elements of old discourses. This is not to deny the element of creativity in Jevons, or any other innovating thinker, but we should investigate closely (and contextually) how such creativity actually operates. It will then become apparent, I suspect, that the creation of a totally novel idea (a Cartesian 'zero hour', so to speak) does not happen that often. Creativity often means recombining elements of existing discourses in original ways and giving the resultant theory a particular new twist. In such a way, ideas change their meaning without being thought up out of the blue. To avoid any misunderstanding, I agree with Bevir that the formation of a new idea cannot be reduced to its contextual antecedents; but I think that his critique of contextualism is far too crude and fails to do justice to the existing practice of historians of ideas.

5. A fifth and last criticism follows from the above. I do not agree with Bevir's criticism of the particular variety of contextualism embodied in the work of John Pocock. Bevir writes: 'if, like Pocock, contextualists believe in various coexisting and overlapping paradigms, they will tend to crush out personal identity. When they tell the history of their putative social languages, they will tend to introduce some utterances by one individual in the story of one language, and other utterances by the same individual in the story of another language, but at no point will they feel a proper need to bring these different utterances together to show the individual held a coherent web of beliefs' (Bevir 1999: pp. 216-17).

   I believe that this criticism is unjustified, because Pocock's 'deconstruction' of authors and texts is to a large extent historically justified. Once again, let it be clear that I'm not at all disputing Bevir's point that an individual seeks to produce a coherent web of ideas, and that the historian ought to appreciate and take seriously such efforts at consistency. There is, however, a great deal of difference between the (modest and defensible) thesis that authors strive after consistency, and the (bold and less sustainable) thesis that they actually achieve it.

   In my view, Pocock is an extremely subtle historian of ideas who operates precisely in the discursive space between those two alternatives: authors seriously seek to achieve consistency, but they do not quite accomplish what they attempt. What we meet with in actual history is a continuous movement between attempted theoretical closure and the multiplicity of discourses engendered by the multifarious intellectual and other experiences individuals have to cope with. Moreover, the failure of a coherent closure is, more often than not, already implicit in the intellectual 'starting position' of an author, for most authors begin their writing career with a heterogeneous set of beliefs and languages: what Bevir rather neutrally calls their 'web of ideas' is in reality fraught with tensions and contradictions. Authors often cherish incompatible beliefs, and, instead of discarding all but one, they spend an intellectual life in trying out new combinations, often adding new ideas, without ever getting quite rid of the original contradictions. Pocock's approach, then, is precisely the one that succeeds in giving a historically plausible account of what happens in such cases. It can explain the long-term persistence of specific discourses or 'languages' in the thought of individual authors, and in many cases across generations as well; and second, it is able to give a non-deterministic, open-ended account of this process in the medium of time. In short, Pocock depicts intellectual innovation as it really is: a process of never-ending bricolage.

IV. Conclusion

Bevir seeks to provide the historian of ideas with a logic (or 'grammar'), a normative account of how intellectual history ought to be done. Even though there are numerous valuable insights in his book from which the historian may profit, his overall approach is too rationalistic and individualist to serve as a comprehensive 'logic' for the history of ideas. Moreover, his treatment of contextualist (and other) currents in the history of ideas fails to do justice to the actual practice of intellectual historians.

References

Bevir, Mark (1999), The Logic of the History of Ideas (Cambridge: Cambridge University Press).

Giddens, Anthony (1979) Central Problems in Social Theory (London: Macmillan).

Hutchison, T. W. (1972) 'The "Marginal Revolution" and the Decline and Fall of English Classical Political Economy', History of Political Economy 4, 442-68.

Roll, Eric (1966) A History of Economic Thought (London: Faber & Faber).

Schumpeter, Joseph A. (1954) History of Economic Analysis (New York: Oxford University Press).

Schumpeter, Joseph A. (1957) Economic Doctrine and Method (London: Allen & Unwin).

Skinner, Quentin (1996) Reason and Rhetoric in the Philosophy of Hobbes (Cambridge: Cambridge University Press).

Winch, Donald (1972) 'Marginalism and the Boundaries of Economic Science', History of Political Economy 4, 325-43.

Discourse index | back to the top

3. Allan Megill - Imagining the History of Ideas

I have read Mark Bevir's The Logic of the History of Ideas with great interest and some profit. It is impossible in the space of an essay to deal adequately with all of its aspects, for the issues that Bevir addresses require at least a book in response. Necessarily, then, I shall be selective. I shall first take note of some aspects of the book that I find interesting and true; I shall then take note of aspects that I find wrong-headed.

Bevir is at his best in addressing historical explanation. Here I want to applaud several features of his discussion. Firstly, he is non-reductionist in his approach to explanation. Much of his concern in the book is with belief, and with how beliefs might be explained. He insists that 'our beliefs resemble a spherical web, not a pyramid'. By this he means that we should not see the task of explaining a belief as if it were a matter of following a conceptual chain that is 'secured at a single point to a particular type of understanding or to a particular type of theory-laden experience'. He opines, for example, that 'explanations of people's conception of God in terms of their political experiences, or their political values by reference to their economic experiences, or their beliefs about wage structures in terms of their religious experiences all seem inherently implausible'. (Bevir 1999: p. 245) In short, he has a general bias for the deployment of different explanans in different regions of human experience, and for the different explanans being located within the region in question.

One has to add, to be sure, that in particular cases explanations rooted outside the region of the belief to be explained might well prove to be persuasive and interesting. One can readily imagine situations where people's beliefs about God could be persuasively and interestingly explained in terms of their economic situations, or their beliefs about wage structures in terms of their religious beliefs. The point is that there is no reason a priori to grant privilege to one kind of explanans over another kind. Bevir writes: 'because all areas of life ... depend on others, we cannot identify any area of life as authoritative over all others' (Bevir 1999: p. 246). I gloss this claim as follows (which may not be Bevir's way of glossing it). Looking at a religious belief from different points of view, one could find illuminating and justified both an 'internal' explanation of that belief, that is, an explanation in terms of other religious conceptions, and an 'external' explanation of it, in terms of non-religious phenomena. Or, one might find several differing explanations for the belief to be illuminating and justified, although from slightly different perspectives. This is not to say that all explanations are to be accepted; it is only to say that different explanations can coexist in the historical field. (This insight tends to work against Bevir's discussion of objectivity [the subject of his chapter 3]), where the argument, perhaps because it is infected by something of a polemical impulse, is not up to the level of his thinking generally (1).

Secondly, Bevir admirably insists on what he calls 'procedural individualism' in the explanation of beliefs and of change of belief. Consider science, with its 'theories, arguments, and problems published in journals and books'. These remain 'mere marks on pages' unless somebody thinks them: 'A theory is a theory only if it is held by someone' (Bevir 1999: p. 232). Further, 'because any theory that exists in time must be a subjective theory held by particular individuals', any problems that emerge with the theory must be 'subjective dilemmas afflicting the webs of belief of specific individuals'. Accordingly, it is not adequate for the historian to claim that a scientist came to believe Y because the scientist existed in problem situation X, and in problem-situation X it is rational to believe Y. Such an explanation--which evokes a Popperian 'world 3' of objective problem-situations--short-circuits historical explanation. Rather, to explain adequately the scientist's new belief, the historian has to grasp 'the dilemmas held and confronted' by that particular individual at that particular time (Bevir 1999: p. 233).

Bevir's insight accords well with a rule of thumb that clear-thinking historians already know, although they may not always articulate it. The rule of thumb in question is that, all other things being equal, "bottom up" explanations are better than "top down" explanations. (I use here the terminology of Steven Shepard, on whose analysis I rely heavily in this paragraph [Shepard 1999]). That is, an explanation will be the weaker, the less it is able to get down to ground level, close to the action to be explained, and the more it has to rely on general claims. Consider, for example, the question of whether Thomas Jefferson had a sexual relationship with his slave, Sally Hemings. One might want to argue for the existence of such a relationship analogically. That is, one would point to the known fact that hundreds of Virginia slave owners had sex with slaves, and one would use this as evidence that Jefferson, himself a Virginia slave owner, probably had sex with Sally Hemings. But as Shepard points out, the existence of such an analogy (between Virginia slave owners generally and Jefferson in particular) 'might help us to interpret the relationship--that is, to see how it makes sense within the larger context of antebellum Virginia', but it 'would not explain the relationship, in the sense of saying what caused it'. For: 'we are concerned with the beliefs and desires of specific actors: unless we argue that Jefferson himself was aware of such an analogy, and altered his beliefs and desires accordingly, then we cannot make use of the analogy in our explanation' (Shepard 1999: pp. 7-8). It is an elementary point, although historians sometimes forget it -- to their cost. It is good to see that Bevir has got it right.

Thirdly, Bevir offers us an interesting theory as to how change of belief takes place. His theory posits that a person's particular beliefs are all part of a 'seamless, holistic web' (Bevir 1999: 246). Experience continually leads people to adopt new beliefs about the world. Some of these beliefs are inconsistent with the person's existing beliefs, and this leads to what Bevir calls a 'dilemma'. A dilemma, for Bevir, is 'an authoritative understanding that poses a question for one's existing web of beliefs'; it is similar to an 'anomaly' in Thomas S. Kuhn's theory of scientific change (as presented in his Structure of Scientific Revolutions) and to a 'problem' in the work of such other theorists of scientific change as Karl Popper, Larry Laudan, and Stephen Toulmin (Bevir 1999: p. 229). Bevir posits that people want their webs of belief to coincide with their experience of reality, and also that they try to make their webs of belief internally consistent. A 'dilemma' thus constitutes a disturbing element within the web of belief. Where a dilemma will lead cannot be predicted, since neither pure observation (which would give people access to the world as it really is) nor pure deductive reasoning is operative. Rather, there emerges 'an inner, Socratic dialogue of question and answer' that leads, although unpredictably, to a change of belief (Bevir 1999: p. 230). Bevir's more general claim is that 'beliefs change only as a result of reflection or thought' (Bevir 1999: p. 247).

Bevir's theory might be seen as standing between the sort of conception of change of belief that is embodied in Hegel's Lectures on the History of Philosophy, and the sort of conception that is embodied in Michel Foucault's The Order of Things (Hegel 1955 [1833-36]; Foucault 1970 [1966]). In Hegel's History of Philosophy, contradictions that arise within one philosophical system lead, necessarily, to the next philosophical system, in which those contradictions are resolved but in which new contradictions arise. In Foucault's The Order of Things, on the other hand, there is no connection at all between one episteme, or system of thought, and the next. Bevir's theory articulates a moderate middle position. The difficulty is that Bevir slides, subtly but unmistakably, from articulating a theory to maintaining that the theory offers the best approach to the real world (Hegel and Foucault tend to do the same thing: it is a theorist's malady). Bevir's theory posits thinking agents who think consciously about their existences, who think rationally (trying with all their might to overcome contradictions), and who express their views sincerely (otherwise we would have inadequate access to what their beliefs and thinking processes actually are). This is fine. What does not stand up is Bevir's claim that, therefore, historians of ideas (who in any case cannot operate without presuppositions) ought to assume in their work, the priority of consciousness over what is not conscious, of rationality over what is not rational, and of sincerity over what is not sincere (2). This simply will not do. For the historian, the questions as to whether someone thought consciously and rationally and expressed herself sincerely are ones that have to be investigated in each particular case. Prior to investigation, the historian ought to remain agnostic as to whether and to what degree the people he or she is investigating thought consciously and rationally and spoke sincerely.

I now turn to what is weakest in Bevir's book: his conception of the history of ideas, which has little -- one is tempted to say, nothing -- to do with the history of ideas as actually carried out by its practitioners. Bevir's 'history of ideas' also has little to do with various fields overlapping with or adjoining the history of ideas, such as intellectual history, history of philosophy, literary history, history of art, history of science, history of mentalities, and cultural history (3). The complete absence from Bevir's book of any attempt to deal with actual work in the history of ideas is striking. And yet he ends his book with the following claim:

The main purpose of my logic...is to promote a particular way of doing the history of ideas. My principal aim has been to describe how historians of ideas should explain the historical objects they postulate and justify the narratives they tell. (Bevir 1999: p. 318)

What is going on here? One might expect a logic of the history of ideas to be set within the framework of an account, or at least an adequate grasp, of historical research and writing in general and of work in the history of ideas in particular. To be sure, some aspects of Bevir's account might well be taken as adequately describing parts of the practice. Some suggestions are ones that historians of ideas, and historians generally, ought to follow, e.g., his prescribing of procedural individualism in the explanation of change of belief. But many of his suggestions, and his perspective on the history of ideas as a whole, would greatly impair the field if they were generally adopted. Indeed, they would destroy it. What does the history of ideas as described by Bevir, or inferable from his work, have to do with the actual historical field of the history of ideas? One has to say, not much.

But Bevir's book is entitled The Logic of the History of Ideas, and one has to ask why this is the case. Much of what Bevir says is very intelligent, and so it is hard to imagine that he does not know at some level, conscious or unconscious, that there is a major disconnection here. How is the disconnection to be explained? By the time I was halfway through The Logic of the History of Ideas, I had articulated a hypothesis on the matter. I cannot say for sure that the hypothesis is correct. However, even if the hypothesis does not explain Bevir's claims, it does provide a way of interpreting them - that is, of making sense of them (4).

Bevir's 'history of ideas' should not be mistaken for the genre to which such illuminating works as Erich Auerbach's Mimesis, M. H. Abrams's The Mirror and the Lamp, Maurice Mandelbaum's History, Man, and Reason, or Hayden White's Metahistory, not to mention many other works of lesser scope, belong (Auerbach 1953 [1946]; Abrams 1953; Mandelbaum 1971; White 1973). Rather, Bevir's 'history of ideas' is an imagined device that he offers us in the course of carrying out a series of philosophical investigations. Bevirian history of ideas is akin to brains in a vat, in Hilary Putnam's suggestion, 'imagine that all human beings...are brains in a vat' (Putnam 1981: pp. 5-8). The imagined brains have little resemblance to real brains: it is a philosopher speaking here, not a neurologist. Bevirian history of ideas likewise resembles the bat in Thomas Nagel's question, 'What is it like to be a bat?' (Nagel 1979: pp. 165-80). Nagel is a philosopher, not a microchiropterologist. His bat has nothing to do with the hundreds of bats that I used to see, on summer nights, wheeling around the dome of the old State Capitol in Iowa City, Iowa, catching insects attracted by the floodlights. Bevirian history of ideas is not far from John Rawls's postulate of the 'original position', in which people imagine themselves behind a 'veil of ignorance' as to how they might be 'advantaged or disadvantaged by social and natural contingencies' (Rawls 1971: pp. 17-22). Rawls is a philosopher, not God, and can only speculate on the original position.

Like Putnam's vatted brain, Nagel's bat, and Rawls's original position, Bevir's 'history of ideas' is part of a thought experiment--an invented proving ground for theories that the philosopher wishes to elaborate on, develop, and criticize. It is not, and should not be mistaken for, the practice that is carried out by, or ought to be carried out by, actual historians of ideas. On occasion Bevir speaks as if it should be, but this is only because, like all of us, he sometimes nods. Or perhaps he was simply carried away by the enthusiasm of his reasoning. In any case, The Logic of the History of Ideas is not a good guide for doing work in the history of ideas and related fields. It is interesting as a theoretical account of belief and of the explanation of change in belief.

One can make further sense of Bevir's approach if one situates it within the history of philosophy. One feature that characterizes much of analytic philosophy (to which Bevir devotes chapter 1) is the disinclination that many of its practitioners show toward dealing with actually existing objects. In the eighteenth century, philosophy was the all-embracing secular discipline; it included both natural philosophy (= natural science) and moral philosophy (= the human sciences) within its kingdom. But today almost everything that was once part of philosophy now belongs to other, distinct disciplines, which were hatched in philosophy but later moved out of the philosophical aviary. One might see this process as one whereby all specific, concrete, empirical curiosity left philosophy and took up residence in other fields, from physics and chemistry, through psychology, economics, political science, and sociology, to history and anthropology. What was left in philosophy was a purely theoretical curiosity, which needs imaginary examples to exercise itself upon, since the real examples all belong, today, to other disciplines. Hence, in this case, the invention of 'the history of ideas'.

It is a tricky enterprise to offer prescriptions for a discipline or a field. One likely error is the error of naive empiricism. This is the error of arriving at a normative account of the field by adding up supposedly good examples of its practice and deriving one's standards from those works. The naive empiricist unselfconsciously thinks to derive normative judgments, or at least to derive a set of implicit recommendations for practice, from what has been done in the field in the past, as if he or she had managed to show why specifically this set of works ought to be taken as generating normative propositions. One is relieved to see that Bevir has avoided this error -- and it is an error, for it makes tradition determinative for a field. The opposite error is the error of naive theoreticism. This is the error of coming to a field with theories, or theoretical preoccupations, originating from outside the field, applying them to the subject matter of the field, and deriving thereby recommendations for the field. To this one ought to say: beware of strangers bearing gifts. Yet, at the same time, one ought not to reject the theoretical impulse, for it is hard to see how, without some help from theory, the history of ideas or any other field in the human sciences can hold itself critically distant from its environment.

Notes

1. Insofar as historical research and writing are concerned, the so-called 'question of objectivity' is a non-issue. It is pointless to ask, of a historical claim, 'Is it objective?' In fact, clear-thinking historians never ask the question. The objectivity question is always a product either of muddled thinking, of a covert polemical intent, or of some combination of the two. Rather, they ask the question, 'Is this claim true?', which immediately leads to the question: What are the grounds for our believing that this claim, more than other possible claims, is true? Further, a historical work could be full of true claims, but at the same time be unworthy of notice by the community of historians. Consequently, clear-thinking historians generally ask other questions as well, most notably: Is this claim original (or is it something that we know perfectly well already)? And: Is this claim interesting (that is, does knowing that the new claim is true actually change anything important in the historical field, or does it leave everything as it was previously, before the claim was advanced)? Note also that the so-called 'question of objectivity' is complicated by the fact that, without further specification, the term 'objectivity' is close to meaningless, since there exist different senses of objectivity, which are related to each other but which have rather different meanings and implications; on this point see Megill 1994: 1-20.

2. For Bevir's discussion of what one might call 'the three priorities', see Bevir 1999: 142-73, constituting the major part of his chapter 4, 'On Belief'. See especially pp. 142, 151, and 158, where sections dealing with sincerity, the conscious, and the rational respectively begin.

3. For a collection of essays that seek in one way or another to define the history of ideas, see Kelley, ed. (1990). The essays in question are drawn from the Journal of the History of Ideas and range in publication date from 1940 to 1987. To be sure, 'the history of ideas' is broader than the JHI tradition; for example, the field includes a philosophically-oriented branch that is rather different from the JHI variety, which values erudition (Gelehrsamkeit) over philosophy. Nor should 'history of ideas' be sharply distinguished from related genres, most notably intellectual history. Nonetheless, the JHI tradition certainly needs to be taken into account in any attempt to legislate for 'the history of ideas'.

4. On the distinction between explanation and interpretation, see Megill 1989. Description, explanation, argument (or justification), and interpretation constitute the four central, and indispensable, tasks of historical research and writing (that they are the four tasks of historiography is discoverable both through a conceptual analysis of what historians must do and through a rational reconstruction of what they actually do do). One problem with Bevir's account is that he tends to assume that the task of history -- or at any rate by far the most important task -- is explanation. But he has no right to make this assumption. Description, explanation, argument or justification, and interpretation all play a role in historical research and writing. There is no justification for giving general priority to one task over the others; everything depends on the context and purposes of the specific work of history in question.

References

Abrams, M. H. (1953) The Mirror and the Lamp: Romantic Theory and the Critical Tradition (New York: Oxford University Press).

Auerbach, Erich (1953) [1946] Mimesis: The Representation of Reality in Western Literature, trans. Willard R. Trask, Princeton: Princeton University Press.

Bevir, Mark (1999) The Logic of the History of Ideas (Cambridge: Cambridge University Press).

Foucault, Michel (1970) [1966] The Order of Things, trans. anon. (New York: Random House).

Hegel, G. W. F. (1955) [1833-36] Lectures on the History of Philosophy. trans. E. S. Haldane and Frances H. Simson (3 vols; London: Routledge and Kegan Paul).

Kelley, Donald R., ed. (1990) The History of Ideas: Canon and Variations (Rochester, NY: University of Rochester Press).

Kuhn, T. S. (1970) The Structure of Scientific Revolutions (2nd rev. enl. ed. (1st ed. 1962), Chicago: University of Chicago Press).

Mandelbaum, Maurice (1971) History, Man, and Reason: A Study in Nineteenth-Century Thought (Baltimore: Johns Hopkins University Press).

Megill, Allan (1989) 'Recounting the Past: 'Description,' Explanation, and Narrative in Historiography,' American Historical Review 94, 627-53.

Megill, Allan (1994) 'Four Senses of Objectivity', in Allan Megill (ed.) Rethinking Objectivity (Durham, NC: Duke University Press), pp. 1-20.

Nagel, Thomas (1979) 'What is it like to be a Bat?', in Thomas Nagel, Mortal Questions (Cambridge: Cambridge University Press), pp. 165-80.

Putnam, Hilary (1981) 'Brains in a Vat', in Hilary Putnam, Reason, Truth, and History (Cambridge: Cambridge University Press), pp. 1-21.

Rawls, John (1971) A Theory of Justice (Cambridge, MA: Harvard University Press).

Shepard, Steven (1999) 'Thomas Jefferson, Sally Hemings, and Inference to the Best Explanation', unpublished paper, Department of History, University of Virginia.

White, Hayden (1973) Metahistory: The Historical Imagination in Nineteenth-Century Europe (Baltimore: Johns Hopkins University Press).

Discourse index | back to the top

4. Kari Palonen - Logic or Rhetoric in the History of Political Thought? Comments on Mark Bevir

I almost stopped reading Mark Bevir's book after the first chapter. The very idea of a 'logic' in the history of ideas remains quite doubtful to me. Fortunately, I continued and realized that the book contains an original perspective on the problematics as well as provides fresh approaches to a number of different topics.

Still, I remain unconvinced about the Bevir's characterization of his research programme as a 'logic'. Indeed, he actually identifies 'logic' with 'grammar': 'to identify the logic of any discipline one has to uncover the forms of reasoning appropriate to it by means of a study of the grammar of the concept operating in it' (Bevir 1999: p. 8, reference without further notice to this volume). But can we 'identify' logic and grammar? Do disciplines have a 'grammar' of their own? What kind of metaphor is 'grammar' in this context?

Both logic and grammar imply the existence of a stable normative system. In my view this Platonic ideal must be rejected. A classical critique of this ideal was presented by Max Weber in his famous Objektivität article. According to him fixing a system with fixed questions and problem areas is already Unsinn (Weber 1973a: pp. 184-5)

Although I side with Mark Bevir in a number of controversial points, notably that of advocating 'procedural individualism', I play here the role of an advocatus diaboli. In particular, I will question Bevir's approach to the history of political thought from a perspectivistic view of knowledge, as proposed by Nietzsche and Weber. My special concern is the intelligibility of conceptual change in political thought.

I. Meaning, conventions and conceptual change

The decisive criterion for the concept of the 'history of ideas' for Mark Bevir is the identification of this concept with the analysis of 'meaning': 'Historians of ideas study relics from the past in order to recover historical meanings. They seek to reconstruct ideas or meanings from the past' (p. 31).

The concept of meaning is, of course, highly polysemic. Bevir proposes to distinguish between the hermeneutic, semantic and linguistic meanings, of which he considers the first to be the proper object of an historical study involving a specific change: 'Any theory reducing hermeneutic meaning to some other type of meaning cannot account for change' (p. 48). I agree with him that, in terms of political thought, these two types of meanings tend to be quite irrelevant.

However the grammar is not a logic-analogous system only. Increasingly also the semantic and linguistic meanings have been politicized, as is the case in the feminist linguistics during the recent decades (for the German debate on the feminist politicization of the grammar see Gorny 1995). Or, to take an example from Bevir (p. 56): I prefer to refer to myself as 'single' rather than as a 'bachelor', for reasons obviously related to the politics of lifestyles.

What does all this mean for Bevir's triad of meanings? An intrusion of the 'hermeneutic meaning' into the semantic or linguistic ones, or the rejection of the triadic distinction as too strict? In any case, we must allow grammatical revisions, for example, to also be used as opportunities for the reconceptualization of meaning. A paradigm of this situation is Koselleck's (1975) thesis on the change of the expression die Geschichte from plural to singular in the late eighteenth-century German and its significance for the revision of the concept of history.

Against the imperialistic claims of an sociological approach, Bevir's critique of contextual determinism is highly justified. The target of his criticism, as it is presented in the following passage, is, however, quite surprising:

Of course, contextualists such as Pocock and conventionalists such as Skinner have written histories that trace conceptual and linguistic change. My point is not that they cannot write such histories. It is that when they do so, their accounts of change cannot be based on the theories of meaning they avow (p. 49).

I shall leave Pocock out of this particular discussion and concentrate instead on the question of whether Quentin Skinner can be considered to be a 'conventionalist'. Skinner's famous article from 1969 already contains a critique of the shortcomings of contextualism, especially in the following passage: 'the fundamental assumption of the contextual methodology, that the ideas of a given text should be understood in terms of its social context, can be shown to be mistaken' (Skinner 1969: p. 43).

Strictly speaking, Bevir extends this critique of social contextualism to the linguistic variant and characterizes Skinner as a conventionalist by quoting this particular passage (p. 88n). Bevir quotes Skinner's article Motives, Intentions and Interpretations of Texts (1972), on which he comments as follows:

Skinner justifies equating intended illocutionary force with actual illocutionary force on the grounds that authors must follow the ruling conventions since they want to be understood. He says: because 'any writer must standardly be engaged in an intended act of communication', therefore 'whatever intentions a given writer may have, they must be conventional intentions to uphold some particular position in argument, to contribute in a particular way to the treatment of some particular theme, and so on' (p. 41).

In his conclusion, Bevir simplifies Skinner's position:

No doubt Skinner is right to suggest that authors standardly perform acts of communication, but this does not allow him to conclude that intentions always must be conventional, or at least it allows him to do so only if he conflates standardly with invariably, in which case his premise is false (p. 42).

Contrary to Bevir's suggestion, we have no reason to assume that Skinner would have 'conflated standardly with invariably'. In his reply to me concerning his renunciation of the assumption of a standard meaning in his turn toward rhetoric (Palonen 1999: p. 48), Quentin Skinner emphasized a shift in his own views: 'It now seems to me, in short, that all attempts to legislate about the 'correct' use of normative vocabularies must be regarded as equally ideological in character' (Skinner 1999: p. 67).

Bevir neglects the conceptual shifts in Skinner's work. Taking the first paraphrase-cum-quotation as a point of departure, I will, moreover, insist that Bevir misses Skinner's point, namely, the rhetorical character of the agent's reliance on conventions, which is something purely instrumental for an 'innovative ideologist' (cf. Skinner 1974) to persuade his or her opponent. Thus, the 'must' in the quotation above is not a logical but a rhetorical 'must', not an obligation but a condition for all wise strategic moves. Using conventions in this way already indicates the agent's distance to those who simply accept them and turn the reference to them into an indirect move in one's own argumentation,

A more interesting argument against Skinner is Bevir's distinction between the notion that 'authors can challenge conventional beliefs and attitudes' and the notion that 'authors can use words unconventionally' (pp. 49-50). Still, I find it quite difficult to comprehend why he thinks that Skinner only discusses the first scenario and not the second. At least from the perspective of rhetorical redescriptions, both are always present (cf. Skinner 1996b: esp. ch. 4).

Furthermore, I am also surprised by Bevir's lack of concern for the key aspect of Skinner's interest in conventions, namely, their role as tacit assumptions in past historical situations.

It is the very impression of familiarity, however, which constitutes the added barrier to understanding. The historians of our past still tend, perhaps in consequence, to be much less aware than the social anthropologists have become about the danger that an application of familiar concepts and conventions may actually be self-defeating if the project is the understanding of the past (Skinner 1970: p. 136).

In historical studies, conventions play the role of implicit and uncritically assumed assumptions, which must be included in the interpretation of past texts. This requires a lot of translation work and imagination concerning 'the past as a foreign country' (op. cit., p. 137). It is in this sense of contingent and implicit yet commonly accepted assumptions that Skinner also speaks of the 'foundations' in his book (Skinner 1978, cf. also Kelley 1970).

Bevir notes in passing (pp. 134-5) Skinner's distinction between the dimensions of meaning and linguistic action. This is probably the point at which the most significant difference between their problematics can be detected. On the role of this distinction Skinner writes in his Reason and Rhetoric in the Philosophy of Hobbes:

Traditional hermeneutics has generally, and often exclusively, concerned with the first of these dimensions [of meaning]; I concentrate very much on the second. I attempt to take seriously the implications of the fact that, as Wittgenstein puts it in the Philosophical Investigations. 'Words are deeds' (Skinner 1996b: p. 8).

I would, thus, count Bevir among the proponents of 'traditional hermeneutics'. When disregarding the dimension of linguistic action they, in a sense, continue the Platonic quest to keep ideas pure from action and politics. Skinner's view by no means denies the dimension of meaning, but instead represents it in terms of the perspective of linguistic action. It is from this perspective that I want to interpret the following passage:

there is a sense in which we may need to understand why certain proposition has been put forward if we wish to understand the proposition itself. We may need to see it not just as a proposition but as a move in argument. To understand it, we may need to grasp why it seemed appropriate to make just that move, and hence to issue just that utterance (Skinner 1996a: p. 146).

In other words, I will question Bevir's claim that Skinner is unable to justify his studies on conceptual changes. The point is that their primary object does not concern 'meaning', but rather meaning is to be made intelligible in terms of linguistic action

II. Criteria of objectivity

Mark Bevir also searches for a third possibility between 'the extremes of objectivism and skepticism' (p. 79). I am again struck to find that Quentin Skinner is included among the objectivists and that his approach of historical interpretation is characterized as 'a logic of discovery' (p. 82), something hardly compatible with his later rhetorical work. Similarly, I am unconvinced that we should reject 'a playful nihilism' (p. 79) or Richard Rorty's critique of 'a neutral, mind-independent work' (p. 111) in the name of objectivity.

The unsurpassed position in all discussions on objectivity can still to be found in Max Weber's famous essay from 1904. My thesis is that Weber achieves a position between objectivism and skepticism in the sense sought by Bevir. From the Weberian perspective Bevir's attempt to make the positive and universal criteria for objectivity is, nevertheless, too substantialistic.

Recent scholarship has accentuated the significance of Nietzsche for Weber's work. A significant aspect of this can be seen in Weber's perspectivistic view of knowledge (cf. e.g. Peukert 1989, Palonen 1998). Perspectivism allows Weber to transcend scepticism without falling back on the unrealistic claims of which objectivism consists. A science of culture presupposes a one-sided emphasis of perspectives. For Weber the idea of an objective world acts only as a limit, which, in all its richness and variety, cannot be exhausted by any conceptual means (Weber 1973a: pp. 170-71).

'Objectivity' is possible only in the struggle between the competing perspectives, for which there should be ample opportunity. This struggle does not imply the acceptance of any relativism between given standpoints or a static shift of positions between definite 'knowledge parties', but it is related to the continuous chances of conceptual revision. In a famous passage of the essay Weber insists on the 'eternal youth of the historical disciplines', due to either the shift in problematics or the revision of conceptual constructions (Weber 1973a: p. 206).

All this concerns the type of criteria necessary for conceptual change and for the revision of knowledge in general. Bevir's ideal criteria are: accuracy, comprehensiveness, consistency, progressiveness, truthfulness and openness (p. 106). Offering such a list already implies a 'positive' claim on objectivity, which subsequently limits both perspectivism and the chances of historical revisions of knowledge. For example, the criterion of 'progressiveness' presupposes the question 'in which respect?': 'progress' of knowledge in some sense may easily be turned into an obstacle in the possibility of conceptual revision. Similarly, Bevir's consensual view on the 'generally accepted facts' (p. 100) can be questioned: the consensus about facts may actually be a result of a 'Chinese stationariness' (Weber 1973a, p. 184, following J. S. Mill). The point of a new perspective can consist of a provocation against the 'facts' of all other existing perspectives.

The 'eternal youth' of the historical disciplines also concerns, above all, the criteria of assessment, which cannot exist independently from perspectives and which themselves are, in principle, subject to radical revision. Instead of general and 'positive' criteria for the assessment of knowledge, the chances of 'objectivity' lie in keeping the competition between alternatives as open and as severe as possible. In this respect, in the Weberian perspective, the practitioners of the human sciences act within a situation similar to that of politics.

III. Explaining historical change

Mark Bevir views explanation and understanding as two separate tasks, as does Weber in his Kategorien article of 1913 (1973b). I think Bevir's programmatically anti-scientistic model of explanation is an original one that deserves closer inspection.

Bevir works within a framework of an interconnected web of beliefs, and he searches for a model of explanation to be applied to changing beliefs: 'Historians should explain beliefs by reference to traditions and dilemmas' (176). For Bevir, 'traditions refer to the initial human condition of being thrown into a world already inhabited not only by people, but also by their definite beliefs'. Speaking of traditions in this sense of an 'initial influence on people' (p. 201) and as being 'contingent entities people produce by their own activities' (p. 203) neither involves a value of order as such nor an obstacle to change, but is compatible with procedural individualism. 'Individuals are agents capable both of modifying traditions and of migrating across traditions' (p. 197).

Bevir's view avoids the pitfall obvious in the jargon of the '-isms' or 'languages', as if they were ready-made packages modifiable only in their margins. His point is not to name the traditions, but rather to understand them in a more formal sense. Traditions 'provide only a synchronic explanation of people's beliefs' (p. 214) but cannot make changes in the beliefs intelligible. What Bevir says about the conceptual change is interesting. His starting point in this respect is the following formula:

conceptual change occurs because all of us are individual agents who reflect on the traditions we inherit in the light of our own experiences and thereby alter these traditions according with our own reasoning (p. 225).

Where are the concepts in this formula? Do the qualities of the concepts themselves and their relationships to other concepts not play a role in conceptual change? It seems to me that Bevir is liable to miss the significance of the role played by concepts as 'pivots' of language (to quote Koselleck 1996: p. 64) and to accept Pocock's (1996) holistic critique against the special significance of concepts in the use of language. Like Koselleck in his reply to Pocock, I would insist that concepts cannot be reduced to the level of the (mere) meaning of terms or beliefs about them, but that they also include the naming and other aspects of the conceptualization of a phenomena. Or, to put it differently, concepts form a bundle of questions rather than supply definite answers.

The key to explaining changes in beliefs as proposed by Bevir is the concept of a dilemma. Resolving a dilemma is a matter of consistency within the individuals' 'webs of belief' (p. 230), which can be dealt with in several different manners but must be 'resolved' either by reconciliation with old beliefs or by changing them (p. 235).

For Bevir, consistency thus appears as a normative requirement and 'changes of belief come about as a result of an inner, Socratic dialogue of question and answer' (230). The Sophist alternative (cf. Cassin 1995) could question the necessity of this transfer of the 'logical' duality to the process of thought. Why not retain 'inconsistencies'? Do they not keep thought open to several options all of them worth being followed more closely? Are dilemmas, ambiguities, paradoxes, internal conflicts, et cetera, not important elements of any thought and worth closer attention? Why to try to 'resolve' them instead of provoking their introduction? Bevir's pathos of consistency seems to exclude or marginalize subversive or system-dissolving chances of dealing with dilemmas.

I think this 'systemic' bias is connected with a Platonic tendency in the search for 'logic'. The rhetorical alternative of arguing in utramque partem, trying to make, à la Protagoras, 'the weaker case stronger', are sacrificed to the logical claim of consistency. I think the 'weaker cases' are worth rescuing. Then, the history of thought would become less a philosophical and more a rhetorical and political enterprise.

IV. Logic or politics in the history of ideas?

Mark Bevir's 'logic of the history of ideas' is based on the Socratic model of self-deliberation. What I wonder is, why he entirely fails to mention the relation of the changing beliefs to the beliefs of others. His variant of the history of ideas seems to bracket the questions of persuasion and dissuasion as bases for changing one's beliefs. In a formal sense, this may be due to the fact that the final decision to change is the individual's own an idea with which I agree against kinds of 'causal explanations' based on 'influence' etc. Still, I think that the exclusion of the rhetorical and political dimensions from dealing with dilemmas and traditions are to be held responsible for the impression that Bevir's 'logic' to end right at the point at which interesting questions begin to become visible.

It is in this sense that I consider his programme of the history of ideas to be related to the older variant of the discipline. The dirty world of politics was kept out of the world of 'ideas', even when 'politics' was said to be the topic. The 'revisionist' programme was once formulated by Quentin Skinner:

For I take it that the political life itself sets the main problems for the political theorist, causing a certain range of issues to appear problematic, and a corresponding range of questions to become the leading subjects of debate (Skinner 1978 I: p. xi).

With Skinner I consider politics to be important for the present discussion in two respects, the first concerning the thematization of questions and the second consider the role of politics in the explanation of the history of thought.

Bevir's view that holds 'beliefs' to be the key unit in the history of thought tends to miss the constitutive role of the question-formation, stressed already by Collingwood (1939). Bevir's denial that this applies to 'logic' seems to signify a reappraisal of the Lovejoyan style of the 'history of an idea', in which the shifting standpoints were analyzed in relation to a more or less ready-made 'agenda' of problems. My specific argument against this practice is the insight that during the twentieth century the character of political conflicts has largely shifted from standpoints to the struggles between the different profiles of questions or agendas. The division of governmental departments no longer allows for a common agenda shared by all parties. This constellation has been replaced by one in which political agents of all kinds compete in the invention of new topics and in getting their own profiles on the political agenda. The 'standpoints' may often look quite similar, although their significance for different agents might in fact vary to a great extent (for example the internet tests for 'choosing' an electoral candidate still tend to presuppose the presence of a common agenda).

Politics is not to be considered as a separate sphere but as an aspect of all human actions. In this sense I consider politics to be relevant for the explanation of all history of thought, independent of the subject matter. Even if the history of thought concerns meaning, I think that it cannot be properly explained without invoking action and politics in particular. The history of ideas should also be analyzed as a kind of 'theory politics'. Shifting one's beliefs can hardly be separated from others' beliefs or from the struggles on the topics on the agenda. All this has implications for politics, both in the intra-academic and in the wider sense. Thus, I do not find politics as something merely contextual but as an inherent part of the study of the history of thought, as viewed from the perspective of action.

V. System construction and system dissolution

In The Logic of the History of Ideas Mark Bevir has constructed a programme all his own for the study of the history of thought, and this is by no means a minor achievement.

I would like to close this commentary with an anecdote that is repeated in all of the stories about the young Sartre. In the late 1920s Jean-Paul Sartre and his petit camarade Raymond Aron used to discuss daily matters of philosophy. The curious thing about their discussion was that it was always Sartre who constructed a theory and Aron who repudiated it. This was an ongoing game of theoretical cat and mouse between Sartre and Aron, and although Aron challenged his theories, Sartre was always prepared to present Aron with a new one when they met. Here, I, myself an old Sartre scholar, have played the role of Aron in my criticism of Mark Bevir's book and, to some extent, toward system builders in general. Let us hope that Mark Bevir continues his work as a system builder and that he, like Sartre, constructs a new system in his next book.

References

Bevir, Mark (1999) The Logic of the History of Ideas (Cambridge: Cambridge University Press).

Cassin, Barbara (1995) L'effet sophistique (Paris: Gallimard).

Collingwood, R.G. (1939) An Autobiography (Oxford: Clarendon Press 1978).

Gorny, Hildegard (1995) 'Die feministische Sprachkritik in der Bundesrepublik', in Georg Stötzel & Martin Wengeler (Hg) Kontroverse Begriffe (Berlin: de Gruyter), pp. 517-62

Kelley, Donald R. (1970) Foundations of Modern Historical Scholarship. Language, Law and History in the French Renaissance (New York: Columbia University Press).

Koselleck, Reinhart (1975) 'Geschichte', Chapters. I, V-VII, in Geschichtliche Grundbegriffe (Stuttgart: Klett), Bd. II, pp. 593-5, 647-718.

Koselleck, Reinhart (1996) 'A Response to Comment on the Geschichtliche Grundbegriffe', in Hartmut Lehmann & Melvin Richter (eds) The Meaning of Historical Terms and Concepts. New Studies on Begriffsgeschichte (Washington: German Historical Institute), pp. 59-70.

Palonen, Kari (1998) Das 'Webersche Moment'. Zur Kontingenz des Politischen (Wiesbaden: Westdeutscher Verlag).

Palonen, Kari (1999) 'Rhetorical and Temporal Perspectives on Conceptual Change', Finnish Yearbook of Political Thought 3, 41-49.

Pocock, J.G.A. (1996) 'Concepts and Discourses: A Difference in Culture?', in Hartmut Lehmann & Melvin Richter (eds) The Meaning of Historical Terms and Concepts. New Studies on Begriffsgeschichte (Washington: German Historical Institute), pp. 47-58.

Skinner, Quentin (1969) 'Meaning and Understanding in the History of Ideas', History and Theory 8, 3-53.

Skinner, Quentin (1970) 'Conventions and the Understanding of Speech Acts', The Philosophical Quarterly 20, 118-38.

Skinner, Quentin (1972) 'Motives, Intentions and Interpretations of Texts', New Literary History 3, 393-408.

Skinner, Quentin (1974) 'Some Problems in the Analysis of Political Thought and Action', Political Theory 2, 277-303.

Skinner, Quentin (1978) Foundations of Modern Political Thought. Vol. 1. The Renaissance. (Cambridge: Cambridge University Press).

Skinner, Quentin (1996a) 'From Hume's Intentions to Deconstruction and Back', Journal of Political Philosophy 4, 142-54.

Skinner, Quentin (1996b) Reason and Rhetoric in the Philosophy of Hobbes (Cambridge, Cambridge University Press).

Skinner, Quentin (1999) 'Rhetoric and Conceptual Change', Finnish Yearbook of Political Thought 3, 60-73.

Weber, Max (1973a) [1904] 'Die 'Objektivität' sozialwissenschaftlicher und sozialpolitischer Erkenntnis'. Gesammelte Aufsätze zur Wissenschaftslehre (Tübingen: Mohr), pp. 146-214.

Weber, Max (1973b) [1913] 'Über einige Kategorien der verstehenden Soziologie'. Gesammelte Aufsätze zur Wissenschaftslehre (Tübingen: Mohr), pp. 427-74.

Discourse index | back to the top

5. Frank Ankersmit - Comments on Bevir's The Logic of the History of Ideas

I wish to begin with expressing my profound respect for Professor Bevir's achievement. He has written a book that deserves to be read by all historical theorists and that undoubtedly is an important contribution to contemporary debate in historical theory. I do not doubt that his book will be widely discussed and that it will prove to be a milestone in contemporary historical thought. All the more so since it suggests how best to overcome the present impasse in historical theory. Contemporary historical theory is torn apart between two extremes: on the one hand there is the extremism of empiricists such as McCullagh, Gorman or Murphey, whereas on the other hand we find the extremism of some of Hayden White's disciples and for whom the link between the historians' text and the past itself has become desperately thin. Bevir opts for a prudent juste milieu between - as he calls it - 'the extremes of objectivism and scepticism' (p. 79); and a flexible and tolerant anti-foundationalism is his guide in his effort. The results of this liberal and un-dogmatic epistemology are most rewarding. To mention two examples, I wholly agree with Professor Bevir's rejection of the exaggerated and pathetic scepticism that one will only too often find in contemporary historical theory. Next, I fully agree with his comparativist view of historical truth as expounded in the section of chapter 2 on 'objectivity through comparison': historical truth is not a matter of a correspondence between historical writing on the one hand and past reality on the other, but requires us to compare a set of rival representation of the past with each other and of seeing, next, which is the most satisfactory one of the set. In my book on narrative logic of some twenty years ago I came to exactly the same conclusion - and on the basis of much the same argument as Bevir's (Ankersmit 1983: pp. 235 - 47. See especially p. 244: 'we can only determine the relative merits of a narratio by comparing it to other narratios').

But though I could go on with singing the praises of Bevir's book, this is not why we are gathered here today and I shall, therefore, in agreement with the purpose of this assembly now turn to a critical discussion of its main claims. Bevir's book mainly is a book about how we should conceive of the meaning of the great texts that we have inherited from the past. We are naturally inclined to identify a text's meaning with the intentions of its author. What a text means is what its author intended it to mean - so we tend to believe. And many hermeneuticists have attempted to justify this common sense intuition. The main purpose of Bevir's book is to defend this hermeneutic tradition against a number of authors from the recent past - such as Barthes, Derrida, Gadamer, Foucault, Skinner or Pocock - questioning this intuition ('it is, therefore, the common sense view that the meaning of a given utterance derives from the intentions of its author, although the exact nature of an intention remains to be determined', Bevir 1999: p. 38).

Bevir's main culprits are Pocock and Skinner. Pocock and Skinner downplay authorial intention by arguing that, respectively, context and linguistic convention, hence, not authorial intention, determine meaning. In the second chapter - which is the most important one of the book - Bevir attacks Pocock's contextualism and Skinner's conventionalism on the basis of a theory of textual meaning. The main claim of this theory is that we must distinguish between semantic and hermeneutic meaning. Semantic meaning is, roughly, what an utterance means by the meaning of the words that have been used for making it, hence what would make an utterance true or untrue; whereas hermeneutic meaning should be identified with the speaker's intentions. He then criticizes the contextualists and the conventionalists for being exclusively interested in semantic meaning and not in hermeneutic meaning.

This distinction between semantic and hermeneutic meaning seems to me both legitimate and helpful. Even more so, I do not doubt that the distinction abundantly makes clear where Bevir and Pocock disagree - though I am less sure here about Skinner. There is a Collingwoodian background shared by both Bevir and Skinner suggesting that their disagreement might well be a matter of philosophical technique rather than of their respective 'authorial intentions' - if I may use here the term that is sub judice in the present context. Bevir then goes on defend the case of hermeneutic against semantic meaning; and he does so by 1) identifying hermeneutic meaning with authorial intention (Bevir 1999: p. 51); and 2) by claiming that it is the main task of the intellectual historian to find out about the intentions of the author of a text. Against this background Bevir defines his own position that he describes as 'weak intentionalism'. And about this 'weak intentionalism' he writes 1) that it 'equates hermeneutic meanings with the meaning utterances have for particular individuals, whether they be authors or readers' (p. 54; see also note 1 below) and 2) that 'weak intentions are the final intentions authors have as they make their utterances' (p. 70; see also note 2 below). And, lastly, in the name of this 'weak intentionalism' the battle is fought against Derrida, Foucault, Pocock, Skinner, Stanley Fish or Eagleton.

Now, though I agree with Bevir that these authors sometimes carelessly and irresponsibly defended their case, I am still not convinced that Bevir's criticism is sufficient justification for his main claim that the historian of ideas should look for the weak intentions of the authors of texts. For there are two more arguments, one theoretical and the other methodological, against intentionalism, whether 'weak' or 'strong', and that Bevir did not consider. Let me begin with the theoretical argument. Bevir thesis is, roughly, that meaning is intention. But it is not immediately clear why this should be so. We could imagine a machine mixing letters and words in an arbitrary way. Most of this will be nonsense, of course. Nevertheless, it is to be expected that now and then a meaningful utterance will be produced. We could select the production of the machine accordingly by means of the following procedure: if an utterance is meaningless, then we put it on a pile on our left; is it meaningful, we put it on a pile our right. The fact that no difficulties are to be expected from this procedure suggests that there is nothing contradictory about the notion of a meaningful utterance not expressing an intention. And this shows that meaning does not presuppose intention in any philosophically relevant sense.

It should be added, furthermore, that this example is less absurd and unrealistic than we might at first sight believe. This will become clear if we think of how lawyers interpret the law. Lawyers will ordinarily not ask for the intentions of the legislator; the only question that interests them is how the law's text applies to a certain case, and how it can, or has in the past been made to apply to similar cases. Or, to put it metaphorically, the trajectory from the legislator's intention to the text is of no significance to them; all legal debate should be situated on the trajectory from text to application. So, from the perspective of the participants in a legal debate it would makes no difference whether some machine, or an assembly of human beings with actual intentions, had produced the law's text: they only have take into account the text as it is. Moreover, we should recall that, in opposition to scholarly debate in the humanities, the interpretation of the law sometimes may truly be a matter of life and death. So when hermeneutic debate may really have far-reaching social consequences, when it may directly affect our life and that of our fellow-citizens and when it is not merely a matter of academic disagreement, intention has to go.

But perhaps Professor Bevir would now riposte that this could not possibly be an argument against his thesis since he is discussing the kind of text that the past has left us. And these are not the productions of the kind of machine (or of the legislative quasi-machine) I was imagining just now, but of intelligent individual human beings. His thesis could then amount to either a strong or a weak claim. The strong claim would be that intention is simply what we mean when speaking of the meaning of a text. But this strong claim could be refuted on the basis of examples given by Bevir himself, such as Mrs Malaprop saying (in a play by Sheridan) 'a nice derangement of epitaphs' when what she probably intended to say was 'a nice arrangement of epithets' (pp. 45 ff). So that leaves us with the weaker claim of Professor Bevir's 'weak intentionalism'. According to this claim 1) intention and meaning are logically equivalent, but 2) as is shown by the Mrs Malaprop example, what is expressed by a speaker or writer need not always coincide with a speaker's or writer's actual intentions. However, in terms of the distinction between semantic and hermeneutic meaning the problem raised by 2) can be adequately dealt with: for unlike semantic meaning, hermeneutic meaning always gives us the meaning of an expression by correctly identifying the speaker's intentions. If so, it follows that, unlike semantic meaning, hermeneutic meaning is equivalent with intention because it expresses intention. And this, then, is how Professor Bevir argues his claim that intention will give us the text's (hermeneutic) meaning.

Let us now concentrate on hermeneutic meaning exclusively and ask ourselves what is the nature of the relationship between (hermeneutic) meaning and intention. Professor Bevir postulates the equivalence of both - and this is precisely why he believes that intentions will give us access to meaning. But, if I am not mistaken, the equivalence of intention and hermeneutic meaning is presupposed by him rather than argued for. First, it is easy to see that there could not be a logical equivalence of (hermeneutic) intention and meaning. For statements about a speaker's intentions, such as 'S intends to say p' are not logically equivalent with statements about the meaning of p. Professor Bevir has probably lost this from sight because of an, admittedly, most unfortunate ambiguity in the word 'meaning'. For compare 1) 'did you really mean to say p?' to 2) 'what did you mean by saying p?': in 1) 'meaning' is equivalent with 'having an intention', whereas in 2) the meaning of a statement is at stake. And these are, obviously, different things since an answer to question 1 does not automatically imply that we now also have an answer to question 2. Nevertheless, we often tend to forget about this difference, since in many contexts our puzzlement about how to interpret a speaker's words is adequately solved by the answer to question 1; but in such cases our puzzlement was a puzzlement not about the meaning of what S said but about why he said it. And, indeed, if we forget about this difference in meaning of the word 'mean' as exemplified by either 1 and 2, we may feel tempted (with Professor Bevir) to equate intention and (hermeneutic) meaning.

But, surely, Professor Bevir might now wish to point out that there can be no textual, or hermeneutic meaning without people intending to mean something when speaking or writing. And if, then, the relationship between intentions and meaning could not to be a logical one - as we have just seen - perhaps we should settle for a causal one instead. The suggestion would then be that intentions explain meaning; and that this is why we should rely upon intentions in order to understand meaning. More specifically, as was suggested a moment ago already, our puzzlement about meaning often is puzzlement about why or how somebody could say or write something under certain historical circumstances. And is this not the kind of problem that we often encounter in the practice of the history of ideas? For example, it might be argued that it was Hobbes's intention to combat anarchy by means of an absolutist theory of the state, or of sovereignty and that this is why he wrote the Leviathan. So invoking Hobbes's intentions here would explain why we have such a book as the Leviathan at all: no intention, no book.

But would this work, as Professor Bevir wants to have it? More specifically, can intentions effectively explain or clarify textual meaning? Answering this question will require us, in the second place, to consider a methodological problem occasioned by Professor Bevir's main claim. As I suggested a moment ago, his argument puts a premium on seeing intentions as (the explanatory) causes of meanings (if you want to explain meaning, look at the author's (weak) intentions). But this picture is at odds with the actual practice of the history of ideas. For authorial intention ordinarily is an epistemological redundancy in the practice of historical writing. This can be explained as follows. Suppose two historians A and B try to discover the (weak) intentions of Hobbes when writing his Leviathan. Suppose, furthermore, that A and B proposed different accounts of Hobbes's intentions - as ordinarily will be the case. In that case historian A might argue: 'Hobbes's intentions were x, as you will see if you take into account what he wrote on pages ABC', whereas historian B might now riposte by saying: 'no, Hobbes's intentions were y if you take into account what he wrote on pages DEF' etc. Now, what to do in such a situation? The natural, but naive answer would be that we should now compare both accounts with Hobbes's (weak) intentions and ascertain, next, which of the two agrees with Hobbes's intentions. But, unfortunately, this is easier said than done. For where do we find Hobbes's intentions? Hobbes's text is singularly unhelpful here; for we do not have both the text of the Leviathan and, apart from that, some postscript in which Hobbes explicitly stated what his intentions were when writing his book. (And if we had such a postscript, it would occasion exactly the same kind of problem as had been raised by the text itself.) All we have, instead, is, first, just Hobbes's text itself and, next, a number of interpretations of the text, taking us to either pages ABC, DEF, GHI, et cetera, for finding out about Hobbes's authorial intentions. So, an appeal to Hobbes's intentions is decisive in no stage of the discussion between A and B.

All this demonstrates that an appeal to Hobbes's authorial intentions is nowhere an ingredient in the process of coming to an optimal understanding of Hobbes's texts (as is is suggested by Bevir's model where intentions explain meanings). The notion of authorial intention only comes in when we have read a specific, let's say X's, book on Hobbes and that we found particularly convincing. And then we may feel tempted to exclamations such as: 'after painstakingly having worked my way through a whole library of books on the Leviathan, X's book finally has shown me what Hobbes's weak intentions really were when writing it'. But this is merely paying a nice compliment to a certain interpretation that we find very compelling and, emphatically, not the sudden discovery of a long looked for correspondence between a historical account (X's text) with part of past reality (i.e. Hobbes's weak intentions).

Hence, the notion of authorial intention is much like Wittgenstein's wheel in the machine and about which he observes: 'a wheel that can be turned though nothing else moves with it, is not part of the mechanism' (Wittgenstein 1974: section 271). And so it is here as well. The notion of authorial intention, whether weak or strong, is such a 'wheel' that is not part of the interpretative 'mechanism' itself. It has no other function than to praise interpretations that we find particularly convincing, but is never part of the argument we use for the discovery and the justification of our interpretative beliefs or convictions. The notion is an epistemological redundancy or, to be more precise, an ontological hypostatization of this redundancy.

I should add that this precisely is the view to which Bevir's comparativist account of historical truth should have committed him. For the main idea of his comparativism is that all we have is a text and many different interpretations of this text, and we decide about the merits of these interpretations by comparing them to each other, and not by measuring their interpretative success by comparing them to the author's intention lying somewhere like a hidden essence in the interpreted text. So I would suggest that precisely Bevir's comparativism will require him to abandon epistemological and metaphysical redundancies such as authorial intention. These redundancies badly need to be shorn off by Occam's razor.

But I should add a few qualifications to this claim. In the first place, it certainly is not my wish to deny the existence of such things as intentions. Neither would I question the legitimacy of the quest for intentions in certain areas of historical writing. For example, in diplomatic or military history historians will often have to find out about the intentions of kings, statesmen and generals. And in such cases the historian may be as successful in identifying the intentions and motives of these historical agents as we are in our daily life when we have to understand the actions of our friends and colleagues. It is not difficult to explain the difference. For in the domain of human action we ordinarily have at our disposal many different means for realizing a certain intention. This is what makes the participation in social life often into such a fascinating experience and why we often ask ourselves why someone opted for a certain means for achieving a certain goal or intention and not some other. In sum, here the relationship between intention and means is contingent and therefore a legitimate object of empirical investigation. But this is different with texts: there is a logical relationship between what we mean to say (intention) on the one hand, and the linguistic means (text) we use for achieving this intention on the other. What we mean is what we mean by saying or by writing something - and not something apart from what we say or write (which does not exclude, of course, that it may sometimes be very difficult to establish meaning and that the quest for meaning may even require us to disregard what a text literally says). And then intention is less independent from the means used for achieving it than is ordinarily the case in the domain of human action. Means will then give us a more direct and reliable access to the past than intentions (3). Nevertheless, I would not want to doubt that we can legitimately ascribe to Hobbes what Bevir defines as 'weak intention' and that we could try to find out about these intentions, if we should wish to. My claim merely is that the function of the notion will in the practice of the history of ideas just as much a hopeless redundancy as when we would inform the scientist that he should find out the 'Truth about nature'.

But I would be prepared to go even one step further and claim that the notion is not merely a redundancy, but a dangerous redundancy as well. It may, under certain circumstances, make us look in the wrong direction. This can best be elucidated by means of an example. But before proceeding further, I would like to make a more general remark about examples in Bevir's book - or rather about the lack of them. In my opinion one of the regrettable things of Bevir's book is the near to complete absence of well-chosen and illuminating examples drawn from the (recent) history of the history of ideas. In fact, I could well imagine someone who has read Bevir's book from cover to cover and even then still has not the foggiest idea of what kind of discipline the history of ideas actually is, nor of what kind of topic it deals with, nor of what kind of problem the intellectual historian is likely to encounter, nor of what is presently discussed in this discipline (4). He may well still be wholly in the dark about how the history of ideas should be demarcated from the history of mentalities on the one hand and the history of science and of philosophy on the other. Moreover, he will still be ignorant of whether, in Professor Bevir's opinion, the history ideas is identical with intellectual history or not, and he will undoubtedly be surprised by Bevir's perfunctory, two sentence dismissal of Arthur Lovejoy (pp. 201, 202), who actually founded the discipline and coined the term 'the history of ideas'.

But I return to my own example. I recently wrote an essay on rococo ornamentation. Most often ornamentation is used for demarcating the three-dimensional space we are living in from the two-dimensional space of a painting or a picture. This is, mainly, what we have ornamental picture frames for. However, rococo ornamentation as devised by Juste-Aurèle Meissonnier in his Livre des ornements et dessines (1734) does precisely the opposite. The rococo ornament does not conduct us out of daily reality into the illusionist reality of the painting, but moves in the opposite direction, from the realm of representation to represented reality itself. Now, in an example like this one, a quest for authorial intention would lead you nowhere. For what is sufficient already is simply to discover this reversal of how ornament relates to three-dimensional reality. Whether this shift has actually been Meissonnier's (weak) intention is a subsidiary question, a question that is of merely biographical and not of any historical significance. And, to put it bluntly, I could not care less.

Moreover, the recommendation to look for Meissonnier's weak intentions would put the cart before the horse. For we do not begin by asking ourselves what intentions Meissonnier had with his engravings in order to discover, as the result of such an inquiry, this shift in how ornament relates to reality. For how could we operationalize the quest for intentions, so that we might hit upon this reversal in the relationship between ornament and reality? I simply have no idea, and I would challenge Professor Bevir to show me in what way intentions might be helpful in an enquiry like this one, as long as we do not yet know about this reversal. For in cases like these, historical inquiry follows the opposite route. We begin by carefully studying the engravings made by Meissonnier's predecessors such as Bérain, Marot, Watteau et cetera, next, we may discover certain formal similarities in them, and then, if we are in luck, we will suddenly recognize where Meissonnier's engravings so strikingly and crucially differ from them. And, then, indeed, only then we can ask ourselves whether the shift was actually intended by Meissonnier, or not. But, once again, this question is of so little historical significance that most historians will not even bother to investigate it. It would address no real historical issue and remain merely anecdotal.

And so it is in the history of ideas. One reads the great texts we inherited from the past and then we may see certain slight, but meaningful, shifts in the use of terms like freedom or sovereignty, or in the intellectual structures of theorists such as Hobbes, Kant or Hegel. This is what intellectual historians mainly do, this is the really difficult part of their job, and authorial intention is wholly tangential to it. This is also where, in my opinion, the true genius lies of a Foucault, Pocock or of a Koselleck.

If, moreover, their methods differ so dramatically from those of Professor Bevir's 'procedural individualism', one wonders whether he would require us to condemn Foucault's Les mots et les choses or Pocock's The Machiavellian moment to the dustbin. I am sufficiently confident in Professor Bevir's openmindedness to expect him to be horrified by any such suggestion. But if so, how is the immense success of the Pococks and the Foucaults to be explained? Bevir's historical theory is sui generis unable to do so. All this should remind us of the old wisdom in historical theory that in case of an outright conflict between historical practice and historical theory it is always the latter that ought to give way. No historical practice has ever foundered on historical theory - and this is as it should be. For historical theory should help us understand historical practice and not condemn it.

I come to a conclusion. In my undergraduate days as a philosophy student my teachers always told me to look at a philosopher's examples if you wish to get an inkling of what they had in mind when writing their books and of what is right and wrong in them. Indeed, I do think that Bevir's examples are quite illustrative of both the strengths and weaknesses of his theory. So let us consider the 'Hallelujah lass' example to which Bevir returns again and again. This example shares with other examples such as 'two is the number of my dogs', the example of the lady unaware of the correct meaning of the word 'arthritis' (p. 63) or the Sheridan example of Mrs Malaprop an ambiguity of meaning. In terms of his opposition between 'semantic' and 'hermeneutic meaning' Bevir explains how the ambiguity can be solved and hermeneutic meaning can be established. This is the good part of his theory. But the bad part is an implicit suggestion about the kind of hermeneutic ambiguity we encounter when dealing with the great texts we inherited from the past. In a way one can argue that these great texts - or 'works' as Bevir would probably prefer to say - are 'ambiguous': after all they permit of many different interpretations. But they do so in a different way than is suggested by Bevir's examples. In Bevir's examples it is as if we have a universe of already existing, fixed meanings, such as 'what a superb solution to the dilemma', 'there is a female member of the Salvation army', or 'there is a zealous Protestant committed to social work' (p. 40). Next, somebody exclaims 'Hallelujah lass', and then the hearer of this exclamation needs to find out what item on the list of already existing meanings the utterance should be linked to, by taking into account the circumstances under which it was made.

But this is not how it is in history. There you do not have at your disposal this list of already existing meanings. There you have to create, so to say, both the list of meanings and to decide to what meaning on the list of meanings a certain text, or work, corresponds. Putting it this way already suggests what the situation in actual historical practice will be like. Historians will feel little inclined to make, first, an exhaustive list of all the meanings a text may possibly have in order to establish, next, which item on the list is the most attractive candidate. The procedure would render historical writing impossible. So what they ordinarily do is to read a text from the past, ponder it, and then they propose a new item to be added to the already existing list of meanings that one may give to the text. And, needless to say, they will offer a rational justification of why they think that this new candidate is the most plausible one. I emphasize, furthermore, that the new candidate really is a new candidate. So where actual historical practice differs from the picture suggested by Bevir's examples is that historical practice is always prepared to recognize what I might paradoxically describe as 'the newness of the past'. Or, to put it in a less paradoxical way, historians are always prepared to discover something 'new' in the past in the sense that they are aware that the past is fundamentally indifferent from the present - and that this is precisely what we have history for. 'The past is a foreign country', to use David Lowenthal's famous metaphor; the past is a strange and alien world and it is the historian's task to fathom the profound and marvelous depths of its radical strangeness. This he can only do when being completely open to the 'newness' of past meaning and when avoiding as much as possible the infringement upon the past of an already existing list of meanings (5).

This, lastly, is where my real disagreement with Professor Bevir lies. I have little sympathy for his attempt to re-introduce the vocabulary of intentions in historical theory. I believe that the notion of intentions is next to useless for a satisfactory understanding of the history of ideas. But this merely is a technical issue. Though I worry about how it might blind Bevir's readers to the achievements of historians such as Pocock or Foucault. Worse is that his reliance on analytical philosophy of language may introduce into historical theory the world picture that is implicit in the writings by Wittgenstein, Ryle, Austin or Strawson. This is the world of daily life, of the conversation with our friends and colleagues and of the kind of simple and commonsense statements we make about everyday experience. And this is a world that is wholly alien to our encounter with the foreign country of the past.

There is nothing wrong with analytical philosophy. On the contrary, we may well hope that its love of precise and accurate argument will liberate us from the gross absurdities abounding in much of contemporary historical theory. But we must never be guilty of forcing the world of the past into the mould of the world that is presupposed by analytical philosophy. This is what Professor Bevir has done in his book and what I find objectionable in it. Instead, we should appeal to all those most refined instruments that were developed by analytical philosophy of language, to the rigour of its argument and to its unparalleled intellectual sophistication in order to come to a better understanding of history. And this will require us not to apply analytical philosophy of language to historical writing but to add a wholly new dimension to analytical philosophy. Analytical philosophy is not an intellectual gold mine for historical theory - it is the other way round - historical theory is an intellectual gold mine for analytical philosophy. An improved awareness of what happens in the practice of historical writing might well provoke a revolution in analytical philosophy without its equal since the days of Wittgenstein, Austin or Strawson.

'The chief business of twentieth century philosophy is to reckon with twentieth century history', thus Collingwood some sixty years ago (Collingwood 1970: p. 79). Indeed, the problem of how a complex text may account for a complex (past) reality has never been put on the agenda of contemporary philosophy. And if philosophers were willing to take seriously this problem - a problem that is of far more practical interest and significance than the often esoteric problems that they so eagerly debate - it is philosophy of history that they should turn to.

Notes

1. I find Bevir's explicit unwillingness to distinguish here between what an utterance might mean to either its author or to its reader a bit puzzling, since most of hermeneutics precisely has to to do with the question of whether this gap between author and reader can be bridged and, if so, how.

2. No mention here of the reader of the text - see the previous note for this.

3. Intentions would, therefore, only be of interest to the historian of ideas in so far as the question arises why a an author has relied upon a specific set of textual means - instead of some potential alternative candidate - for expressing a specific intention. Hence, in a situation where intention and the means for expressing intention are seen as being independent from each other and where, thus, the relationship between the two has become the main topic of investigation. A rhetorical analysis is then required, since rhetorics focusses on this relationship.

4. Bevir's examples are usually derived from the history of science or, rather, from the standard stock of examples that is in use amongst philosophers of science. A reliance on this type of example will undoubtedly result in an over-intellectualization of the history of ideas. For scientific rationality is less prominent in the history of ideas than in the history of science. This over-intellectualization has left its unmistakable traces in Bevir's chapter on 'Diachronic explanation': Bevir here presents the history of ideas as a chain of responses to ever new dilemmas. And, surely, this is how one could look at the history of physics. But what 'dilemma' was involved in the re-discovery of the category of the sublime in the 17th and 18th centuries, or in the creation of the realist novel by Balzac, Flaubert and the Goncourts?

5. This is also why I would totally disagree with Bevir's assertion that Pocock's and Foucault's views on the writing of history testify of a lack of historical sense (see pp. 49, 60). No historical theory - either from the past or the present - so openly recognizes the depth and the range of historical change than their thesis of the evolution of intellectual and political vocabularies.

References

Ankersmit, F.R. (1983) Narrative logic. A semantic analysis of the historian's language (The Hague/Boston: Kluwer, Martinus Nijhoff).

Bevir, M. (1999) The logic of the history of ideas (Cambridge: Cambridge University Press).

Collingwood, R.G. (1970) An autobiography (Oxford: Oxford University Press).

Wittgenstein, L. (1974) Philosophical Investigations (Oxford: Oxford University Press).

March 2001

Bevir's response to Munslow

Author's Response (2)

Discourse index | back to the top