Interpreting Texts Adrian Blau

Download 130.64 Kb.
Size130.64 Kb.
  1   2   3
Interpreting Texts
Adrian Blau

Senior Lecturer in Politics

Department of Political Economy

King’s College London
Draft chapter for Adrian Blau, ed., Methods in Analytical Political Theory

(Cambridge University Press, 2016).

 N.B. As with other chapters in this book, text in bold is ‘how-to’ advice for students. 

— DRAFT 1: 16 July 2015 —

Comments and criticisms welcome,

but by 1 August if at all possible!

(The manuscript is to be submitted on 1 Sept.)

8263 words, plus references

1. Introduction: meanings and understandings

There are three main ways of interpreting texts, based on three kinds of understanding and three kinds of meaning: what authors mean, what the ideas mean, and what one or both of the above mean to the reader. Roughly, these kinds of meaning are empirical, philosophical and aesthetic, respectively.

The last of these is primarily the province of literature departments. Of course, for those of us in other departments, what texts mean to us can still influence what we study. Rousseau has made me cry, because of the beauty and passion of his writing. Habermas has made me cry, for other reasons. But what these texts mean to us is not usually our intellectual focus and I discuss it no further.

By contrast, the second kind of meaning – what the ideas mean – is very important for us. Not that you would know it from our methodological literature, which mainly emphasizes the first kind of meaning: what authors mean. Yet both kinds of meaning matter when interpreting texts, because they involve different kinds of understanding. If you read J.S. Mill and understand exactly what he meant by the words he used, you have understood something very significant. But you understand his writing better if you also spot his ambiguities, contradictions, successes and failures. Unfortunately, the best methodological writings about textual interpretation – those by Quentin Skinner (2002a) – say almost nothing about the second kind of meaning and understanding, and imply that the first can be achieved on its own.

I seek to connect these two kinds of meaning and these two kinds of understanding. A piece of research typically prioritizes one of the two, but almost always deals with both. They are not alternatives: we usually need the first to find the second, and we often use the second to find the first. That fundamental point has not, I believe, been made in previous methodological discussions.

Equally unfortunately, most methodological writing gives the wrong impression by talking about different ‘approaches’ or ‘schools of thought’, like the Cambridge School, Straussianism, Marxism, and so on. These categories have some value and I cover them below. But everyone reading this chapter knows that mental categories influence thoughts and actions. Past authors could not think what we think, and we cannot think what future people will think. We have all learned new distinctions that let us see things differently: our previous categories were holding us back.

So, it is not outlandish to suggest that the categories with which many of us think about methods of interpretation have constrained our thinking. Most troubling is when commentators imply that these approaches are very different without adding that there are principles of good interpretation which apply to all of us (see especially Rorty 1984: 49; Dunn 1996: 19; Ball 2004: 19; Richter 2009: 7-11; Schulz and Weiss 2010: 284, 287-8). This chapter’s main aim is to make explicit these universal principles of good interpretation. You will find relevant principles in every section, even for categories you might think do not fit you.

Three brief caveats. First, my examples are mostly historical but the principles apply to all texts. For example, Seana Shiffrin’s (2004: 1644-62) careful probing of Rawls’s ambiguous discussion of racial equality reads like someone navigating tricky passages in Hobbes or Locke.

Second, my examples mainly come from well-known Western authors, like Plato and Machiavelli. But there are good reasons to study other thinkers (Stuurman 2000: 152-65; see also the chapters in this volume by Ackerly and Bajpai on comparative political theory, and by Leader Maynard on ideological analysis).

Third, and most important, although I cover both empirical interpretation (e.g. what Locke meant by ‘trust’) and philosophical interpretation (e.g. how well Aristotle’s arguments work), and although I link the two more than other commentators do, most of my how-to guidance involves empirical interpretation. Other chapters in this volume will help more for readers primarily interested in philosophical interpretation (especially the chapters by Brownlee and Stemplowska on thought experiments, Knight on reflective equilibrium, and Frazer on moral sentimentalism). But you should still read this chapter to the extent that you want to get historical authors right.

Section 2 summarizes the Cambridge School of interpretation: despite its crucially important focus on history and context, there are other secrets to its success, and pitfalls we must all beware. Section 3 addresses Begriffsgeschichte, conceptual history, and genealogy, which combine Cambridge-School interpretation with conceptual comparison. Section 4 tackles philosophical approaches, a category largely missing from previous accounts despite being very common. Philosophical analysis even matters for historians, I argue. Section 5 considers reconstruction, which is sometimes seen as a purely philosophical approach, but which actually we all do. Section 6 questions the usefulness of treating perspectives like feminism or Straussianism as if they are ‘approaches’. They provide hypotheses and distinctions which help us see things that other scholars overlook, but there are far more ‘approaches’ than we usually think. Nor is ‘reading between the lines’ exclusively associated with Straussian interpretation, section 7 argues. Section 8 then stands back and offers more general principles of good interpretation that apply to all of us – whatever categories, approaches or schools of thought we identify with.

2. The Cambridge School

The increasingly misnamed Cambridge School – most of its key figures have now left Cambridge – has dominated our methodological literature for half a century. The Cambridge School has essentially won the battle. Although aspects of its proponents’ arguments remain controversial, their core claim is widely accepted: place texts in their historical contexts. Even consulting such historical research without doing it oneself makes misinterpretation less likely.

The Cambridge School arose in the late 1940s and came to prominence in the 1960s, with the theoretical writings and substantive interpretations of writers like Peter Laslett, John Pocock, Quentin Skinner, and John Dunn (Pocock 2006: 37-9). Given diversity within and between these writers and their followers (Boucher 1985: 151-272), I focus mainly on Skinner, who I regard as the supreme methodologist in our field. While I have learned much from his methodological writings, I have actually learned more from his substantive interpretations, which I find even more methodologically impressive. By contrast, Mark Bevir (1999: 40-50, 327) only tackles Skinner’s methodological writings and seems to me to miss part of Skinner’s brilliance (see also Stuurman 2000: 319; Skinner 2002a: 178-9). Treating Skinner as a practitioner, not just a theorist, lets me sidestep his speech-act theorizing, which is essentially separate to contextual analysis (Hutton 2014: 927) and which I address elsewhere (Blau 2014).

Although Skinner does not write this, and although most commentators emphasize his contextualism, the foundation of Skinner’s success is actually close textual analysis. Indeed, this is how he teaches history of political thought to graduate students, I understand. The first principle of Cambridge-School analysis, and all sensible textual interpretation, is thus: read an author’s texts carefully. Dipping in, which we inevitably do with some authors, is dangerous.

Reading texts carefully means that we should try to read passages in their textual context. Consider Hobbes’s comment that ‘the Thoughts, are to the Desires, as Scouts, and Spies, to range abroad, and find the way to the things desired’ (Hobbes 1651, chapter 8 paragraph 16, p. 53). Read out of context, this sounds as if Hobbes is implying that reason is the slave of the passions. Read in the context of the chapter as a whole, Hobbes is actually not discussing reason at all (Blau 2015c: section 4.2).

We should thus read an author’s texts widely, to avoid overlooking important ideas elsewhere in the text and in other texts, including ‘non-political’ texts. Rousseau thought that Machiavelli’s Prince only praised Cesare Borgia to indicate insincerity and perhaps esoteric intent (The Social Contract book 3 chapter 6 paragraph 5, p. 95 – henceforth abbreviated in the form TSC 3.6.5: 95), but Machiavelli was praising Borgia much earlier (Skinner 1981: 9-12). Machiavelli’s play Clizia gives us different perspectives on the relationship between women and virtú than we get from the Prince and Discourses (Zuckert 2004). Hobbes’s Leviathan may be his most advanced text, but it makes more sense alongside other Hobbes texts: for example, his account of individual deliberation is more detailed elsewhere (e.g. Hobbes 1656). In practice, though, we cannot always read as much of an author’s output as we would like.

Reading an author’s texts widely may tempt us to exaggerate authors’ coherence (Skinner 2002a: 67-72). This is a huge danger. When an author seems unclear or contradictory, philosophers often try to make this coherent, as with Rawls’s (2007: 199-200) interpretation of Rousseau on amour propre. Philosophically, it is fine to construct a Rousseauian ‘system’ of ideas that makes more sense than Rousseau. After all, most things make more sense than Rousseau. But empirically, this may or may not be what Rousseau had in mind: authors make mistakes or change their minds. You never know when to read different ideas/texts into each other and when they are not consistent. Consider both hypotheses.

Textual analysis is never enough: a key Cambridge-School contribution is to place texts in linguistic context by reading other texts from the similar time/place, or by reading the work of scholars who have done this. This can let us understand words we no longer use, like ‘dehortation’, and helps us spot ‘false friends’ – words which look familiar but which once meant something different, like ‘police’, ‘pleasant’, ‘prejudice’, ‘pretend’, or ‘politic’.

Placing texts in their linguistic context may not solve our problems. For example, when Hobbes discusses the ‘dictates’ of reason’ (e.g. De Cive chapter 3 section 19, p. 51), this could mean that reason makes us do something, i.e. acts as a dictator, or that reason tells/dictates to us what to do but lets us decide. Unfortunately, both senses were used in Hobbes’s day, by Hooker and by Donne respectively (Blau 2015c: section 5). I doubt that contextual analysis will resolve this: ultimately we must think through Hobbes’s comments philosophically, to infer how strong reason is in his account. Combining empirical and philosophical analysis is emphasized throughout this chapter, but the omission of this point in Cambridge-School methodological writings is a major gap, or as academics like to say, a major ‘lacuna’.

Placing texts in their linguistic context also helps us infer intentions and indicates originality. Machiavelli uses ‘fortuna’ conventionally but ‘virtú’ unconventionally – dramatically undercutting the orthodoxies of his day (Skinner 1981: 24-31, 34-47). If we only read the ‘canon’ of great thinkers, we will miss part of what makes these texts so ground-breaking. This should be remembered when we hear claims that the Cambridge School had ‘a hugely negative impact’ by making the study the study of past thinkers ‘merely antiquarian’, even ‘frivolous’ (Kelly 2006: 48). But understanding how authors subtly knifed their lesser known contemporaries can make these texts wonderfully exciting. Moreover, as we will see, ‘merely antiquarian’ scholarship can help contemporary normative analysis.

We should thus place texts in their intellectual contexts – political, philosophical, religious, and so on. Consider Rousseau’s claim that a man can be forced to obey the law and still be free (TSC 1.7.8: 53). Rousseau’s justification is unclear. Helena Rosenblatt suggests that Rousseau, like many opponents of the Genevan government, implies a traditional Christian/Calvinist view of freedom. ‘Just like abiding by God’s will makes men free in Christian thought, abiding by the general will makes citizens free in Rousseau’s thought’ (Rosenblatt 1997: 255-6; also 246-7). This explanation, whereby adhering to the general will thereby makes one free, fits part of what Rousseau writes and is surely relevant. But Rousseau immediately goes on to describe the relation in instrumental terms: forcing someone to obey the general will makes them free by guaranteeing them from personal dependence. This is not captured by Rosenblatt’s explanation. I suspect that Rosenblatt has been too quick to assume that the contextual parallel provides the whole answer. Again, we need both contextual and philosophical analysis here. Unfortunately, half a century of Cambridge-School methodological theorizing has not addressed this issue.

Moreover, historical parallels can be coincidental. According to Richard Tuck, Hobbes was responding to a form of scepticism (1993: 285-7, 293-8, 304-7, 316). Most commentators, though, explain the apparent parallels differently (e.g. Zagorin 1993: 512-8). No evidence, including contextual evidence, is conclusive: the same evidence can always be read differently. Sometimes the problem is not an incorrect context but inattention to other relevant contexts: thus Skinner’s work on liberty has arguably neglected the contexts of political economy and theology, which imply somewhat different conclusions (Whatmore 2006: 121-5).

Such points have not been adequately theorized by Cambridge School advocates, who have trumpeted the value of historical interpretation of texts without saying much about how to do it well (Green 2015: 436). The starting-point, as section 8 argues, is to place hypotheses, inference and evidence centre-stage. Historians who doubt this should consider what happens when one ‘knows’ one’s answer is right and just looks for evidence that fits it, as with Leo Strauss (Blau 2012).

Contextualization can even help with recent authors. We understand Rawls better by placing him in his political and philosophical contexts (see, respectively, Forrester 2014; Woolf 2013). Cambridge-School advocates have been curiously quick to read and criticize Rawls without placing him in context. Some do not even read him carefully (see e.g. the criticisms of Raymond Geuss by Sagar 2014). But in truth we can understand much of Rawls without contextualizing him – just as with some passages in historical texts. Contextual analysis, usually helps, but it is not always necessary, and often insufficient.

3. Begriffsgeschichte, conceptual history, and genealogy

Begriffsgeschichte (in English: the ‘history of concepts’) is often associated with Reinhard Koselleck, who co-led the Geschichtliche Grundbegriffe (in English: Basic Concepts in History), a multi-volume, 7000-page analysis of over 100 social and political concepts, published in German between 1972 and 1997. This enterprise is so big, requiring so many years and so many authors, that I do not cover it here. (For more information, see Richter 1996. For much more detail on the Geschichtliche Grundbegriffe in particular and Begriffsgeschichte more generally, see Richter 1995, especially pp. 124-42 on similarities and differences between Begriffsgeschichte and Cambridge-School analysis.)

I focus instead on the smaller-scale version of Begriffsgeschichte: conceptual history, or genealogy. I treat the terms as equivalent: Skinner now calls his conceptual history of the state (1989: 90-126) a genealogy of the state (2009). (For other kinds of genealogy, see Lane 2012a: 75-82.) Conceptual history has long been practised informally but arrived self-consciously in the Anglo-American mainstream with Terence Ball, James Farr and Russell Hanson’s edited book Political Innovation and Conceptual Change (1989). Each chapter takes a concept (e.g. democracy, patriotism, rights) and examines different conceptions of that concept over time and sometimes place. (For the concept/conception distinction – i.e. a general idea, and particular versions of that idea – see Rawls 1971: 5.) This is interesting and important not only historically and politically but also normatively: we see why some conceptions won out, how different contemporary conceptions often are to earlier ones, and how earlier ones might revive contemporary discussions. You don’t need to use conceptual history to make normative arguments, but be attentive to the normative implications of your interpretations in case you can see an application to a contemporary argument.

One potential pitfall is failing to distinguish a word and an idea, as with Peter Euben’s (1989) conceptual history of corruption. I am not sure that Thucydides’s idea of stasis involves corruption – an inevitable problem when one deals with different languages, which is why Skinner sometimes sticks to Anglophone texts (2009: 325). The problem applies even within one language. What Euben says Hobbes says about corruption is actually what Hobbes says about ‘sedition’: his comments on ‘corruption’ are very different (Blau 2009: especially 614-5). Euben’s analysis remains instructive but risks confusing readers. So, be conscious, and if possible explicit, about the extent to which you are focusing on the word and/or the idea.

We can think of conceptual history as having two parts: empirical and conceptual. The empirical part is often Cambridge-School analysis, where as Skinner writes, the primary task is to recover authors’ own understandings (Skinner 2002a: 50). The second stage involves conceptual comparison. In other words, stand back and compare authors’ understandings – ask if Hobbes and Locke understand liberty in the same way, for instance. (See Olsthoorn’s chapter in this volume, and also Rehfeld’s.) Careful conceptual analysis is vital, as with Skinner’s fine-grained genealogy of liberty (2003: 22).

You may even want to apply anachronistic conceptualizations. Anachronism is dangerous and can infect our efforts to recover authors’ meanings (Skinner 2002a: 49-51). But if handled carefully, and preferably not until you have first recovered authors’ meanings, it lets us apply conceptual frameworks that highlight similarities and differences between authors. For example, Rousseau’s Considerations on the Government of Poland wants citizens to elect deputies every six weeks, on an explicit set of instructions, divergence from which would see deputies being executed (Rousseau 1997: chapter 7 paragraphs 14-19, pp. 201-3). This is a great example of the so-called ‘mandate’ or ‘delegate’ conception of representation (Pitkin 1967: 145-7). Rousseau did not use these terms and refused to call this ‘representation’ (TSC 3.15.5-6: 114). But what he says amounts to how we use these terms. So, first try to work out what authors meant, then see how well this fits your own conceptualization or an existing one.

Case selection needs attention. The ideal – taking all possible cases – is impractical, especially in a single chapter or article. Be conscious, and perhaps explicit, about how your case-selection may limit your conclusions, such as whether your claims are restricted to a particular time and place (e.g. Skinner 2009: 325). You may also need to consider words not used, as well as words used, as with Josiah Ober’s analysis of ‘democracy’ and similar terms in ancient Greece (2008: 5, 7).

4. Philosophical analyses

I now address what we might call ‘philosophical’ approaches (following the terminology of e.g. Wokler 2012: 121). Philosophical analysis is extremely common in politics and philosophy departments, and even historians do it. Yet previous categorizations cover it at best partially, or not at all (Rorty 1984; Dunn 1996: 19; Ball 2004; Ball 2011: 49-57; Schulz and Weiss 2010: 284-8; Klosko 2011).

Philosophical approaches need the first kind of meaning mentioned at the start of this paper: what authors mean. But the main focus of philosophical approaches, or at least philosophical parts of other kinds of analysis, is on the second kind of meaning: what the ideas mean/imply. Examples include how well an author’s arguments work, as with A.P. Martinich’s (2005: 80-105) testing of Hobbes’s laws of nature; how certain ideas fit together, as with John Gray’s (1996: 70-85) reconstruction of Mill on happiness; and what we can learn today, as with Skinner on liberty (2003: 24-5).

Note that these scholars are, respectively, a philosopher, a political theorist and a historian. Although Skinner’s early writings avoided normative engagement, even rebuking Strauss for describing Machiavelli as a teacher of evil (1981: 88), more recently he encourages historians to seek contemporary insights from historical texts (e.g. 2002a: viii, 89, 125-7). Just as conceptual history had two essentially different parts (one empirical, one conceptual), so too here: the first stage is Cambridge-School empirical analysis, the second stage involves applying this for contemporary purposes.

And – crucially – philosophical analysis can even help scholars whose main aim is to uncover authors’ meanings and motives. To infer what Hobbes meant by liberty, and why he changed his account, it helped Skinner to think through the philosophical implications of Hobbes’s arguments. Section 2 thus argued that historians may need philosophical analysis to uncover authors’ meanings.

Section 2 also noted that philosophical readings may legitimately diverge from what authors actually thought. Skinner (2003) starts by working out what historical republicans meant, but then, when applying their insights on liberty to contemporary issues, rightly drops their gender assumptions. Gray also begins with a careful reading of Mill, but in seeing how well Mill’s ideas interconnect, is explicitly agnostic about whether this is what Mill had in mind: all the pieces are there and they form a coherent system, but we just do not know how conscious Mill was of this (1996: 70-85). Melissa Lane’s ‘unabashed appropriation’ of Plato maintains ‘the structure of his effort’ but consciously changes many details (2012b: 23, 25). Gregory Kavka (1986) starts by seeking to understand Hobbes in his own terms, then finds problems with Hobbes’s account. This leads Kavka to undertake a ‘reworking’ of Hobbes’s flawed understanding of powers, ‘filling in a critical gap in Hobbes’s argument’ about the state of nature, and revising Hobbes’s arguments in ways that are broadly in tune with his project but which he could not have conceived of. Kavka’s theory is explicitly ‘Hobbesian’, not Hobbes’s (1986: xiv, 3-4, 93-107).

So, be conscious, and preferably explicit, when you diverge from historical authors. For example, Rousseau seems to talk about democracy in two different ways: with and without deliberation (e.g. respectively, TSC 4.2.10: 124 and 2.3.3: 60). Imagine that you want to use Rousseau to defend a deliberative view of democracy, but you are not sure that he preferred this view; perhaps you even suspect that he ultimately preferred the non-deliberative view. Either way, it is fine to use these comments as if they are supporting deliberative democracy, provided you indicate if your Rousseau may not be authentic. Refer to a ‘Rousseauian’ or ‘broadly Rousseauian’ account, for example, and only say ‘Rousseau argues’ or ‘Rousseau believes’ when you are indeed trying to stay true to Rousseau. This will not stop some historians from making dismissive noises about you ‘pillaging the classics’ to find ideas for modern times (Tuck 2007: 69). But thankfully, there is not yet a law against doing this.

Download 130.64 Kb.

Share with your friends:
  1   2   3

The database is protected by copyright © 2022
send message

    Main page