SEARCH

SEARCH BY CITATION

This paper seeks to introduce to the biblical studies community a little-known philosopher of science and of mathematics, Imre Lakatos, with a view to showing that one can profitably apply the insights of his work on the nature of scientific research programmes to the area of evaluating the hermeneutical role of critical thinking in biblical interpretation. The work of Lakatos, as we shall note, has been suggested as a suitable framework for theological thinking by Nancey Murphy. Here I want to propose that in the field of biblical studies, a hermeneutical proposal may be considered in a parallel way to a scientific conjecture which has observable consequences. The advantage of the conceptuality offered by Lakatos, I shall suggest, is that it moves the discussion beyond matters of personal or theological preference and ingrained predilection, towards some criteria which might permit inter-subjective testing and agreement.

After an introduction to the work of Lakatos, I explore some of the key areas of his thought for biblical studies, and then conclude with an example which seeks to show the benefits of this approach: a consideration of the classic theory of Pentateuchal source criticism (JEDP) and some of its strengths and weaknesses. The goal is not to uncover a hitherto neglected essential key to biblical criticism, but is the more modest attempt to suggest that in this case some inter-disciplinary borrowing might shed a little light on some familiar concerns of biblical studies methodology.

IMRE LAKATOS (1922–74)

  1. Top of page
  2. IMRE LAKATOS (1922–74)
  3. TYPES OF LAKATOSIAN RESEARCH PROGRAMMES (LRP'S)
  4. INSIGHTS FROM LAKATOS'S MODEL FOR BIBLICAL STUDIES
  5. AN EXAMPLE: THE JEDP HYPOTHESIS AS A LAKATOSIAN RESEARCH PROGRAMME
  6. LAKATOSIAN HERMENEUTICAL RESEARCH PROGRAMMES IN BIBLICAL STUDIES

Imre Lakatos was a Hungarian-born philosopher of science and mathematics who shared with his London School of Economics colleague Karl Popper the conviction that the academic activity of professional scientists (and mathematicians) must be considered seriously by philosophers, and that only by so doing would philosophy provide genuine insights into scientific knowledge. However, the direction he developed differed from that of Popper, and — if one word generalisations may be permitted — then where Popper focused on the key idea of falsifiability,1 Lakatos's programme centred around the concept of dubitability within a broader framework. In brief: where Popper allows theories to be falsified by data which does not fit, Lakatos seeks ways to adapt ongoing research in the light of new data.

Lakatos looked extensively at the processes which human beings go through in developing their knowledge of a field, and compared deductivist and heuristic approaches to such knowledge. In the former we tend to find formalised presentations of abstract reasoning. In the latter we focus on problems and problem-solving pursuits. His own work centred on ‘the methodology of scientific research programmes’, the title of his most significant paper.2 His most extended example of this approach was actually in the field of mathematics, a text posthumously reconstructed as Proofs and Refutations.3 He also offered preliminary application of his approach to the methodology of historiographical research programmes – an area of obvious interest to biblical scholars although here his thinking was less developed (and has been more criticised).4

It is not my concern to offer an overall description or evaluation of Lakatos's work, most of which lies well beyond the fields of possible relevance to biblical scholars.5 After a brief description of salient features of it, I shall consider the very limited range of theological reflection which has drawn upon it, and then address the particular question of how Lakatos's approach can help to articulate in a constructive way some familiar questions pertaining to approaches to biblical interpretation. A case study is offered by way of illustration of the value of the approach.

I will attempt to clarify all instances of Lakatosian terminology along the way. One additional note: for the purposes of this paper I shall simplify a complex discussion and try to maintain consistent distinction between the following two terms: hermeneutics, which refers to theories of interpretation (or ‘second order’ discussion of interpretation), and biblical interpretation, by which I mean actual interpretations of biblical texts (‘first order’ discussion of interpretation).

TYPES OF LAKATOSIAN RESEARCH PROGRAMMES (LRP'S)

  1. Top of page
  2. IMRE LAKATOS (1922–74)
  3. TYPES OF LAKATOSIAN RESEARCH PROGRAMMES (LRP'S)
  4. INSIGHTS FROM LAKATOS'S MODEL FOR BIBLICAL STUDIES
  5. AN EXAMPLE: THE JEDP HYPOTHESIS AS A LAKATOSIAN RESEARCH PROGRAMME
  6. LAKATOSIAN HERMENEUTICAL RESEARCH PROGRAMMES IN BIBLICAL STUDIES

Lakatos argued (as well as demonstrating to some extent historically) that in scientific and mathematical reasoning we do not generally follow Francis Bacon's model of pure inductive reasoning. Rather we operate by way of a series of what he called ‘bold conjectures’ which serve as trial theorems and which we then attempt to prove. This attempt demonstrates in turn ways in which the original theorem needs revising. In particular he had in mind the misleading way in which mathematical proofs introduce non-intuitive concepts to secure their results when in actual fact these concepts had only arisen historically after the proof as a result of seeing how the proof needed to be tightened up. Concepts thus introduced to deal with irregularities and problems he called ‘proof-generated’.6

Once a conjecture has survived initial revising as a result of attempts to prove it, it takes on the status of a theorem which then serves to guide further research. Theorems acquire a high degree of working reliability as their expected consequences are observed. When these consequences are unobserved, or where there are unexpected consequences observed, then either the theorem is modified, or, in extreme cases, abandoned for a new one. The distinction with Popper is clear here: rather than falsification leading to overthrow, Lakatos allows the ‘research programme’ time to adapt and seek to accommodate the new data, only allowing outright ‘falsification’ as a last resort. The key question is then whether the theorem has explanatory (or predictive) power in its field, or whether it is sustainable only through repeated modification and does not in turn seem to serve a further explanatory purpose. He classified scientific research programmes according to this positive or negative role of the theorems at the heart of the programmes.7

Thus a Lakatosian research programme (LRP) in the sciences is one which seeks to explain or defend a central hypothesis by way of examining its suggested observable consequences. It is progressive if it in turn generates further insights. It is degenerating if it consists in shoring up a programme against consistent problems or attacks and does not in turn generate further new insights.

It will be helpful to explore a little more the structure of a research programme in the sense envisaged by Lakatos. Despite its apparent distance from the normal concerns of biblical studies, the relevance of his approach is I think actually clearest if we look at Lakatos's own formulation of the process gone through in a mathematical proof. This will be of some relevance to biblical criticism in one additional way: it will become clear that one cannot maintain an uncontested notion of ‘absolute truth’ even in mathematics, where formalist philosophies of disembodied number systems have had a hard time of it in recent reflection.8

In mathematics, if not the sciences generally, ‘there is a simple pattern of mathematical discovery … :

  • 1
    Primitive conjecture
  • 2
    Proof (a rough thought-experiment or argument, decomposing the primitive conjecture into subconjectures or lemmas)
  • 3
    ‘Global’ counterexamples (counterexamples to the primitive conjecture) emerge9
  • 4
    Proof re-examined.’ (In other words: a ‘guilty lemma’ is made explicit and then built in to the revised conjecture as a condition of it; and the new conjecture replaces the primitive one, with its focus being what Lakatos called a ‘proof-generated’ concept.)

Amongst various other stages that often follow this process, the most interesting is Lakatos's observation that ‘counterexamples are turned into new examples – new fields of enquiry open up.’10 These new fields can, in general, be best understood when their historical development is known. It is this which allows him to account for the appearance of otherwise non-intuitive concepts which make ‘bold appearances’ in conjectures without apparent rationale to anyone encountering the conjecture unaware of the process of conjecture and refutation behind it.

The promise of such an approach lies in its suggestion that there might be criteria for evaluating the ‘progressive’ nature of a research methodology. As Nicholas Wolterstorff has noted, it is very easy for theological proposals to ‘go nowhere’ and ‘lead to no research programme’ no matter how well intentioned they may be.11 It is not necessary for our purposes to show that Lakatos's own approach is superior to well-known alternatives. The contrasts between Lakatos, Popper and Thomas Kuhn may be left for another day.12 As one simple example of a contrast: unlike Kuhn's comparatively widely-known view of ‘paradigm shifts’ in the philosophy of science,13 Lakatos viewed different explanations (or theorems) as competing paradigms which all at the same time seek ways of accounting for established data and helping to predict new data. In my judgment this is more obviously relevant to the concerns of biblical critics than the Kuhnian notion of successive paradigm shifts, but in any case, my goal is not to say that biblical critics must consider Lakatos's view so much as to suggest that there are some conceptual advantages if they do so.

INSIGHTS FROM LAKATOS'S MODEL FOR BIBLICAL STUDIES

  1. Top of page
  2. IMRE LAKATOS (1922–74)
  3. TYPES OF LAKATOSIAN RESEARCH PROGRAMMES (LRP'S)
  4. INSIGHTS FROM LAKATOS'S MODEL FOR BIBLICAL STUDIES
  5. AN EXAMPLE: THE JEDP HYPOTHESIS AS A LAKATOSIAN RESEARCH PROGRAMME
  6. LAKATOSIAN HERMENEUTICAL RESEARCH PROGRAMMES IN BIBLICAL STUDIES

That Lakatos's approach could be applied to the role of doctrines in theology has been explored by Nancey Murphy in her book Theology in the Age of Scientific Reasoning, which seeks to defend the view that doctrinal theology is a form of empirical science susceptible of a Lakatosian analysis.14 Elsewhere she has written of Lakatos as a ‘postmodern nonrelativist’, as part of her own work attempting to describe a theological methodology which accepts some of the critiques of classic ‘modern’ thought without selling its soul to some of the rhetorical flourishes of ‘continental postmodernism’, to which end she adopts the label of ‘anglo-american postmodernity’.15 The work of Lakatos offers one key resource for her own attempt to defend a view of theology as scientific in a post-foundationalist (if not postmodern) sense.16

Murphy herself notes that ‘many find the application of Lakatos's scientific terminology to theology jarring’ and that Alasdair MacIntyre offers comparable insights in terms more likely to help theologians ‘feel more at home’.17 Perhaps for this reason her proposals have found little explicit uptake in theological discussion, in contrast to the voluminous literature considering MacIntyre's work. In what little discussion there has been, some questions have been asked considering whether she offers a sufficiently ‘scientific’ account of the ‘hard data’ of theology (such as the traditional confessions of the church – which she more or less accepts as they are).18

Murphy's own suggestions concerning the biblical texts as ‘data for theology’ are brief. She does include one example of a Lakatosian analysis: the work of G.B. Caird and Hendrik Berkhof on principalities and powers as providing observable support for a sociopolitical version of the principalities and powers in line with a classic doctrine of the atonement.19 However, scripture here is essentially only playing a role within the analysis of doctrines, and is not considered as a subject in itself. Beyond this brief example I am not aware of further attempts to utilise the framework of Lakatos's ideas in biblical studies.

The lack of follow-up in theological circles of Nancy Murphy's use of Lakatos parallels the general situation in philosophy, even philosophy of science, where Lakatos's own work has not been anywhere near as influential as that of Popper and Kuhn. Of Lakatos himself, Larvor notes ‘The methodology of scientific research programmes has attracted many admirers but few followers’,20 and the problems of adapting his insights are harder in the area of historiographical research programmes, where it is often suggested that he allows the philosophy to run roughshod over the practice of historians, preferring instead a ‘rational reconstruction’ of the history of science which, say critics, does not do justice to the phenomena it seeks to explain.21

Nevertheless, without proposing that Lakatos's work offers a panacea for all matters of hermeneutical complexity or the profound clashing of paradigms in biblical studies, I do want to suggest that there are insights to be gleaned from his understanding of how conjectures and refutations work which could be used to shed some clarifying light on the discipline of biblical studies. Hermeneutical approaches, or frameworks, as they are brought to bear on biblical texts, act in a way more or less congruent with a version of scientific methodology: hypotheses followed by subsequent attempts to verify or reject. Bold conjectures regarding biblical texts are a staple of biblical criticism, but the resultant discussions concerning them often seem to dissolve into little more than statements of preference or the random making and meeting of objections. What Lakatos's methodology of research programmes offers is the chance to evaluate whether such conjectures lead to progressive or degenerating lines of enquiry, and to suggest some criteria which might be operative in such a discussion. I will attempt to illustrate this claim with an example (itself a Lakatosian move!), and then offer some brief pointers to areas where further work might be done.

AN EXAMPLE: THE JEDP HYPOTHESIS AS A LAKATOSIAN RESEARCH PROGRAMME

  1. Top of page
  2. IMRE LAKATOS (1922–74)
  3. TYPES OF LAKATOSIAN RESEARCH PROGRAMMES (LRP'S)
  4. INSIGHTS FROM LAKATOS'S MODEL FOR BIBLICAL STUDIES
  5. AN EXAMPLE: THE JEDP HYPOTHESIS AS A LAKATOSIAN RESEARCH PROGRAMME
  6. LAKATOSIAN HERMENEUTICAL RESEARCH PROGRAMMES IN BIBLICAL STUDIES

In keeping with the spirit of Lakatosian enquiry, I shall first explain why I choose this example. The JEDP hypothesis in Old Testament studies can be considered as analogous to a mathematical theorem: it is a conjecture about the origin and development of the Pentateuch which attempts to account for data which appears contradictory on other accounts. In terms of its historical status as a hypothesis, it is helpful to recognise that it is itself ‘proof generated’ in Lakatos's sense – in other words it is in turn generated by problems with an earlier conjecture, the basic claim that Moses was author of the Pentateuch.22 This ‘primitive conjecture’, in the Lakatosian sense, sought to uphold a traditional ascription of the five books to Moses, as well as to deal with such New Testament data as Jesus' references to the books, teachings and indeed words of Moses when referring to sayings from the first five books (e.g. Matthew 8:4; 19:7; 22:24 and comparable sayings in other gospels).

While it is not impossible to find defences of Mosaic authorship of the Pentateuch today, they hold little sway over the scholarly community, because it has been widely accepted that too much data does not fit the model. The most obvious ‘global counterexample’ (i.e. textual issue which does not fit the theory) is the account of Moses' own death in Deuteronomy 34. Less well known, but nevertheless delightfully tricky for the view of Mosaic authorship, is the description of Moses as ‘the most humble man on the earth’ in Numbers 12:3, which was something of a problem to traditional theories of Mosaic authorship: ‘the self-commendatory nature of the statement occasioned difficulties to older commentators’, as George Gray noted a century ago.23 Such counter-examples are the ‘guilty lemmas’ of Lakatos's approach. Nevertheless, rather than simply seeing such recalcitrant data as sufficient for a Popperian ‘refutation’, one could try to turn these ‘guilty lemmas’ into features of the theory and defend a revised version of it: Moses ‘basically’ wrote the Pentateuch, with some (innocent) exceptions. Of course, such a theory would have had to acknowledge Moses' use of considerable amounts of traditional material in Genesis, though in extremis one might have argued that the material had been revealed (or re-revealed?) at the time he wrote it.

One can see why Mosaic authorship did not survive as a critical hypothesis (or ‘research programme’). The status of what has replaced it is more complicated. In general there are two main lines of argument in defence of something like the JEDP hypothesis: the existence of narrative repetitions in the finished five books, and the conundrums of the occurrence of the divine name. I want to suggest, though, that their status in terms of how Lakatos describes scientific research programmes is different.

Let us begin with the second of these. The crux of the matter occurs, as is well known, in Exodus 6:2–3, with the divine speech: ‘I am Yhwh. I appeared to Abraham, Isaac, and Jacob as El Shaddai, but by my name Yhwh I did not make myself known to them.’ This verse is tricky to square with Gen 15:7; 28:13 and so forth, and indeed all three of the named Patriarchs engage explicitly in one way or another with Yhwh at some point in the text of Genesis. Overall, Gen 4:26 seems to say that people began to call on the name of Yhwh very early on, while Exodus 6:3 seems to suggest that it is newly revealed. While other solutions do exist, 24 the standard view is that separate sources have here been combined: the J source which uses the name ‘Yhwh’, and the E source which uses ‘elohim’. This conjecture, however, turns out not to have enough explanatory power: there are texts which do not fit a J/E conception, and thus two further sources are posited: the priestly source (P), and a Deuteronomic voice which gives us D (represented in the book of Deuteronomy). As the hypothesis grows more complex so more and more ‘guilty lemmas’ are brought to the fore, and incorporated into the hypothesis itself.25 As just one example, early (i.e. 17th and 18th century) versions of the theory dated P early, representing ancient priestly tradition, but in a bold reconceptualisation of the whole approach, Graf dated the law later than the prophets, and P became the final source. As Wellhausen wrote, ‘I learned … that … Graf placed the Law later than the Prophets, and, almost without knowing the reasons for the hypothesis I was prepared to accept it.’26 This is a classic example of an adjustment to a research programme by way of an incorporation of recalcitrant data into a revised version of a theory, and while it might seem as if such an admission as Wellhausen's is self-undermining (a theory posited for no intrinsic reason), in fact adopting a modification to a theory because of its explanatory power is a fair way to proceed, as long as such a modification is then in turn subjected to critical testing.

The case regarding narrative repetition is, in my judgment, less persuasive as a rationale for the theory of written Pentateuchal sources. Too many of the examples do not withstand close scrutiny, not least the obvious problem with what is often adduced as the textbook case of repetition: the stories of the Patriarchs who pass off their wives as sisters for fear of their lives. Versions of this tale occur three times in Genesis (Gen 12, 20 and 26) and, on most accounts, two of these are attributed to J, which seems therefore to be heuristically inadequate to account for the repetition.27 However, if one reconceptualises the phenomenon of repetition not as a generating consideration of the theory, but as an observable consequence of the theory, in Lakatosian terms, then it becomes considerably more plausible. In other words: the point about narrative repetitions does not generate the theory, but it does fit well as an outcome one might expect if indeed J and E have both had their stories woven together. This would in turn require us to ask whether reading narrative repetitions as source-derived is a progressive move: i.e. does it in turn yield new insights which can then be confirmed in other observable consequences? The answer to this question seems to be stuck at ‘possibly’: it is perhaps too easy to resolve narrative tensions by recourse to sources, and those seeking literary unity and coherence have developed an impressive track record in finding reasons to read the relevant texts in other ways.28

What can be concluded from this (excessively) brief survey? I would suggest that the JEDP hypothesis has a certain conceptual plausibility; that it is itself generated by ‘guilty lemmas’ in earlier (simpler) theories about the origins of the Pentateuch; and that is has observable consequences in the field of biblical interpretation which offer considerable reason to suspect that source-critical approaches to the Pentateuch have something significant to offer the biblical critic. To this extent, it operates as a progressive LRP. By way of straightforward contrast, attempts to hold on to Mosaic authorship clearly operate as degenerating LRP's: shoring up the theory with ad hoc modifications which have little or no further explanatory power.29 The point here is that such arguments are often couched in wide-ranging terms with regard to theological persuasion, logical possibility (or at least non-impossibility) and so forth. Lakatos's conceptuality allows us to classify approaches quite simply with respect to their methodological features, and in a fresh way which might help to move some familiar debates further forward.

It may be helpful, by way of further illustration, to consider one case of a scholarly attempt to overthrow the theory, by none other than the authors of a state-of-the-art analysis of Pentateuchal sources themselves: Antony Campbell and Mark O'Brien, in their Rethinking the Pentateuch. Prolegomena to the Theology of Ancient Israel.30 They here express considerable dissatisfaction with the whole project, noting, in Lakatosian terms, that ‘the problems are still there; the Documentary Hypothesis is no longer solving them’,31and thus concluding that the time for protecting the research programme from recalcitrant data is over. To this end they abandon it in favour of an alternative paradigm focused on expansions of short ‘base texts’ which underlie the finished Pentateuch. They thereby dispense altogether with J and P, suggesting too that the work of Noth, Westermann and others had already allowed the conclusion that ‘the need for an Elohist source has faded from the scene’.32 The key shift to conceptualising texts as ‘user bases’ (e.g. for oral performance, allowing narrative development in the subsequent telling) rather than as finished products, may be understood as a Kuhnian paradigm shift, and the rejection of JEDP may be understood as a Popperian refutation, but in fact the methodology of their own work is best grasped in Lakatosian terms: prolonged attempts to harness data to an overarching framework, with the difficulties then in turn generating an alternative conceptuality – much of the book works out hitherto problematic aspects of J and P in terms of the new theory. One should note that the rejection of sources does not lead to a return to Mosaic authorship, and this in turn highlights a key distinction between critiquing JEDP in Lakatosian terms (i.e. attempting to further its explanatory power) and the reactionary response to it on the part of some conservative authors whose only goal in highlighting difficulties was to allow a retreat to the more familiar ground of Mosaic authorship. This would be a contrast, one might suggest, between progressive and degenerating forms of critique within the research programme.

Two further points may be made about the JEDP hypothesis in this perspective. Firstly, it is indeed one rival hypothesis among many, as noted above. Most obviously, it competes against literary analyses which suggest that the finished Pentateuch exhibits too great a coherence to be an edited document. Competing paradigms are a feature of Lakatos's account. Here we must be careful to note too that the different levels at which hermeneutical theories operate can be accounted for in terms of their progressive or degenerating (and indeed their ‘proof-generated’) nature. For example, canonical analysis often suggests that the biblical interpreter should switch paradigms to a different level of enquiry, but this need not (and in practice often does not) invalidate historical-critical enquiry, since it is often dependent upon something like a source analysis in order to trace its hermeneutical stages of development toward a finished product.33 The existence of successful analysis at other levels does not in itself invalidate source criticism – that would be an argument about appropriate goals rather than method – and it would be possible to say, for instance, that something like the JEDP hypothesis could function as a progressive LRP with respect to the status of our knowledge of how the Pentateuch was written, but as a degenerating one with respect to how this helped the final form of the Pentateuch to be understood.

Secondly, having established a kind of ‘internal’ conceptual plausibility to the theorem, this is not the same as saying that the theorem is vindicated, because external factors remain significant. A simple example here would be historical estimations of the nature and extent of writing at different stages in the Old Testament period, a contested issue itself, which has led many scholars to suppose that the Pentateuch's written sources (as against its oral traditions) should be dated quite late.34 The current state of Pentateuchal criticism suffers from all manner of ad hoc estimations about sources, dates and so forth. It will be of use to frame such questions in terms of whether the research programmes they represent are progressive or degenerating.

LAKATOSIAN HERMENEUTICAL RESEARCH PROGRAMMES IN BIBLICAL STUDIES

  1. Top of page
  2. IMRE LAKATOS (1922–74)
  3. TYPES OF LAKATOSIAN RESEARCH PROGRAMMES (LRP'S)
  4. INSIGHTS FROM LAKATOS'S MODEL FOR BIBLICAL STUDIES
  5. AN EXAMPLE: THE JEDP HYPOTHESIS AS A LAKATOSIAN RESEARCH PROGRAMME
  6. LAKATOSIAN HERMENEUTICAL RESEARCH PROGRAMMES IN BIBLICAL STUDIES

I have not attempted to provide anything like a full-scale Lakatosian analysis of Pentateuchal criticism: that would be a massive undertaking. It has been my more modest goal to offer enough of an illustration to show that the workings of hermeneutical theorems in biblical studies can be conceptualised in Lakatosian terms, and to give some examples of how this gives a vocabulary for addressing evaluative questions in somewhat more methodologically robust terms than personal or theological preference, or vague notions of plausibility or implausibility.

Clearly one could pursue such an analysis with any ‘hermeneutical theorem’ in biblical studies. While the JEDP theorem seems self-evidently progressive in the stages where its initial contours and justification are sketched, it becomes notoriously degenerating as its details are pressed to account for ever more tricky observable consequences in the text. Nevertheless that it has a certain degree of solidity might be glimpsed by way of a contrast with another well-known theorem in Old Testament studies: the supposition of a so-called ‘succession narrative’ in 2 Samuel 9–20 and 1 Kings 1–2. This seems to lack anything like the empirical warrant to be sustained, and it does not obviously contribute to the shaping of on-going reflection on the chapters it purports to account for. It bears all the hallmarks of being a degenerating research programme.35 Further ‘theorems’ readily suggest themselves for Lakatosian analysis as research programmes: the existence of a Deuteronomistic History, the questions of Q and gospel sources, gospel audiences, and so on.

I do not claim that the work of Imre Lakatos is an essential component of self-reflection on the part of biblical studies as a critical discipline, but there are sufficient points of contact between Lakatos's consideration of scientific research programmes and the nature and function of critical hypotheses in biblical studies to suggest a fruitful interaction. In this paper I have tried to lay out the framework of Lakatos's approach, indicate how its concerns might be taken up in biblical studies, and show, in theory and by way of example, that there may be gains in clarity and understanding of the processes of scholarly claim and counter-claim if we consider describing interpretative proposals in biblical studies in terms of the methodology of hermeneutical research programmes.

Footnotes
  1. 1 Popper's major work is Karl R. Popper, The Logic of Scientific Discovery (London: Hutchinson, 1959 [1934]), but a similar approach to that of Lakatos can be found in idem, Conjectures and Refutations (London: Routledge and Kegan Paul, 1963).

  2. 2 Imre Lakatos, ‘Falsification and the Methodology of Scientific Research Programmes’, originally in I. Lakatos and A.E. Musgrave (eds.), Criticism and the Growth of Knowledge (Cambridge: CUP, 1970), 91–196; and reprinted in Imre Lakatos, The Methodology of Scientific Research Programs. Philosophical Papers, Volume 1 (eds. J. Worrall & G. Currie; Cambridge: CUP, 1978), 8–101. Worrall and Currie also edited his Mathematics, Science and Epistemology. Philosophical Papers, Volume 2 (CUP, 1978).

  3. 3 Imre Lakatos, Proofs and Refutations. The Logic of Mathematical Discovery (Cambridge: CUP, 1976).

  4. 4 See, in the first instance, his essay ‘History of Science and its Rational Reconstructions’, in Methodology, 102–38, esp 131–36.

  5. 5 Lakatos's own main writings are now available in the two volumes of collected papers and the text on mathematics cited in notes 2–3 above, as well as Imre Lakatos and Paul Feyerabend, For and Against Method. Including Lakatos's Lectures on Scientific Method and the Lakatos-Feyerabend Correspondence (ed. Matteo Motterlini; Chicago: University of Chicago Press, 2000). An excellent overview of his work is offered by Brendan Larvor, Lakatos: An Introduction (London and New York: Routledge, 1998), and a fascinating study locating his philosophy in his Hungarian context is John Kavadny, Imre Lakatos and the Guises of Reason (Durham, NC: Duke University Press, 2001). Note also several focused case studies of wider relevance in K. Gavroglu, Y. Goudaroulis and R. Nicolacopoulos (eds.), Imre Lakatos and Theories of Scientific Change (Boston Studies in the Philosophy of Science 111; Dordrecht, Boston & London: Kluwer Academic, 1989). A strikingly unsympathetic portrait of Lakatos as an ‘irrationalist’ is offered in D.C. Stove, Popper and After: Four Modern Irrationalists (Oxford: Pergamon Press, 1982). Stove's main argument against Lakatos is infact very brief, majoring on terminological issues (9–14), and may well have been misled by the views of Paul Feyerabend, one of his other ‘irrationalists’.

  6. 6 These notions are explored throughout Proofs and Refutations, but see especially 88–92 and 144–54.

  7. 7 Lakatos, ‘Methodology’, 31–47.

  8. 8 The wider setting and ramifications of Lakatos's ‘falllibilist’ philosophy of mathematics are explored in Philip J. Davis and Reuben Hersh, The Mathematical Experience (London: Penguin, 1980), 345–59; cf also Reuben Hersh, What is Mathematics, Really? (London: Vintage, 1997), 208–16. Both of these works foreclose on any uncontested notion of ‘formalism’ in mathematics. A rare attempt to correlate developments in the philosophy of mathematics with models for theological thinking, with particular reference to set-theoretic constructions of infinity, is offered by F. LeRon Shults, Reforming the Doctrine of God (Grand Rapids: Eerdmans, 2005), esp 30–35.

  9. 9 Lakatos, Proofs and Refutations, 127.

  10. 10 Lakatos, Proofs and Refutations, 128.

  11. 11 Nicholas Wolterstorff, Reason within the Bounds of Religion (Grand Rapids: Eerdmans, 2nd ed; 1984), 105–6. He is discussing claims of the kind ‘x needs to be set in the context of God's creation’ as an example.

  12. 12 See e.g. Larvor, Lakatos, 37–46 and 55–57; Ilkka Ninluoto, ‘Corroboration, Verisimilitude, and the Success of Science’, in Gavroglu et al (eds.), Imre Lakatos, 229–43; and Lakatos's own papers on Popper, especially ‘On Popperian Historiography’, in his Mathematics, Science and Epistemology, 201–10.

  13. 13 Thomas Kuhn, The Structure of Scientific Revolutions (2nd enlarged edition, Chicago: University of Chicago Press, 1970).

  14. 14 Nancey Murphy, Theology in the Age of Scientific Reasoning (Ithaca and London: Cornell UP, 1990). See esp 51–87 for her account of Lakatos, and 174–211 for an outline of ‘scientific theology’. Another notable use of Lakatos's insights in theological discussion is by Philip Clayton, Explanation from Physics to Theology: An Essay in Rationality and Religion (New Haven: Yale University Press, 1989) 48–58, who offers a lucid account of Lakatos over against Feyerabend's reading of him as a methodological anarchist. One may also note Gregory R. Peterson, ‘The Scientific Status of Theology: Imre Lakatos, Method and Demarcation’, Perspectives on Science and Christian Faith 50 (1998), 2231.

  15. 15 Nancey Murphy, Anglo-American Postmodernity. Philosophical Perspectives on Science, Religion, and Ethics (Boulder, CO: Westview, 1997), esp 52–55.

  16. 16 Note that Larvor concludes his book with a discussion of Lakatos's relationship to some postmodern French thought, suggesting that ‘Lakatos is cousin rather than brother to these French philosophers’, Lakatos, 108 (cf 107–10). Note that, contra the implications of Stove (n.5 above), Lakatos is absent from recent scientific critiques of postmodernism, such as Alan Sokal and Jean Bricmont's entertaining Fashionable Nonsense. Postmodern Intellectuals' Abuse of Science (New York: Picador, 1998).

  17. 17 Nancey Murphy, Beyond Liberalism and Fundamentalism. How Modern and Postmodern Philosophy set the Theological Agenda (Valley Forge, PA: Trinity Press International, 1996), 103 (cf 100–103).

  18. 18 See especially J. Wentzel van Huyssteen, ‘Is the Postmodernist Always a Postfoundationalist? Nancey Murphy's Lakatosian Model for Theology’, in his Essays in Postfoundationalist Theology (Grand Rapids: Eerdmans, 1997), 73–90, esp 83. Murphy's use of Lakatos is broadly defended in James Wm. McClendon, Jr. and James M. Smith, Convictions. Defusing Religious Relativism (Valley Forge, PA: Trinity Press International, 1994), 140–44.

  19. 19 Murphy, Theology, 178–83.

  20. 20 Larvor, Lakatos, 95.

  21. 21 Larvor offers an excellent discussion of this issue: Lakatos, 99–104; cf Lakatos's own essay cited in n.4 above.

  22. 22 Since a detailed history of Pentateuchal analysis is not my goal here I refer the reader instead to the standard accounts, especially Ernest Nicholson, The Pentateuch in the Twentieth Century. The Legacy of Julius Wellhausen (Oxford: Clarendon Press, 1998) and R.N. Whybray, The Making of the Pentateuch: A Methodological Study (JSOTS 53; Sheffield: JSOT Press, 1987), which basically contests the hypothesis. See also n.25 below.

  23. 23 George Buchanan Gray, Numbers (ICC; Edinburgh: T&T Clark, 1903), 123.

  24. 24 There is a careful review of the issues involved and the options available in R.W.L. Moberly, The Old Testament of the Old Testament. Patriarchal Narratives and Mosaic Yahwism (OBT: Minneapolis: Fortress, 1992), 36–78. His own view is that Genesis was written by a later redactor, who did know the identity of Yhwh, even while that identity was unknown during the events described in Genesis.

  25. 25 An up-to-date analysis which faithfully represents the complexity is Antony F. Campbell and Mark A. O'Brien, Sources of the Pentateuch. Texts, Introductions, Annotations (Minneapolis: Fortress Press, 1993). See n.30 below for a striking update from these authors.

  26. 26 Julius Wellhausen, Prolegomena to the History of Ancient Israel (reprint Eugene, OR: Wipf and Stock, 2003; orig. Meridian Books, 1878), 3.

  27. 27 I have addressed alternative accounts of the nature of repetition in the Pentateuch and elsewhere, including this example, in my ‘The Theological Function of Repetition in the Old Testament Canon’, HBTh 28 (2006), 95–112, subsequent to which I have also come across the thorough analysis of Aulikki Nahkola, Double Narratives in the Old Testament: The Foundations of Method in Biblical Criticism (BZAW 290; New York & Berlin: Walter de Gruyter, 2001), esp 73–114, who argues that ‘double narratives’ have been key to a wide range of reflection on ‘method’ in biblical criticism, even if separate sources can no longer be maintained.

  28. 28 Cf the comments of Whybray, Making of the Pentateuch, 52–53 on the literary approach of Robert Alter, as well as such ‘literary’ readings as Laurence A. Turner, Genesis (Readings: Sheffield: Sheffield Academic Press, 2000).

  29. 29 James Barr offers an exemplary analysis of the doctrine of ‘inerrancy’ in fundamentalism in parallel terms, demonstrating that it involves ad hoc modification of its hermeneutic solely in order to protect the theory, which is exactly, in Lakatosian terms, an example of a degenerating research programme. See James Barr, Fundamentalism (London: SCM, 1977), 40–55. For an illuminating update on Barr's work with regard to fundamentalist hermeneutics see Harriett Harris, Fundamentalism and Evangelicals (Oxford: OUP, 1998) esp 278–324 on ‘Evangelicals and Hermeneutics’.

  30. 30 By Antony F. Campbell, S.J. and Mark A. O'Brien, OP (Louisville, KY: Westminster John Knox Press, 2005) – see n.25 above.

  31. 31 Campbell and O'Brien, Rethinking the Pentateuch, 1.

  32. 32 Campbell and O'Brien, Rethinking the Pentateuch, 10, 21.

  33. 33 This is more obvious in the tradition inspired by J.A. Sanders (e.g. David M. Carr, Reading the Fractures of Genesis: Historical and Literary Approaches (Louisville, KY: Westminster John Knox Press, 1996) but can be seen too in the work of Brevard Childs, such as his attempt to account for the signficance of the canonical redaction of the book of Isaiah from its separate sources (most notably in his Introduction to the Old Testament as Scripture (Philadelphia: Fortress Press, 1979), 311–38).

  34. 34 On writing in Old Testament times see the studies of e.g. David M. Carr, Writing on the Tablet of the Heart: Origins of Scripture and Literature (New York and Oxford: Oxford University Press, 2005) esp 253–72 where he links the consolidation of Scripture to the Hasmonean period; Susan Niditch, Oral World and Written Word: Orality and Literacy in Ancient Israel (London: SPCK, 1997), who argues that the basic conceptuality of the JEDP hypothesis ‘comes from our world and not from that of ancient Israel’ (112); and the various essays gathered in Piotr Bienkowski, Christopher Mee and Elizabeth Slater (eds.), Writing and Ancient Near Eastern Society. Papers in Honour of Alan Millard (LHBOTS 426; London: T&T Clark International, 2005), especially M.C.A. Macdonald, ‘Literacy in an Oral Environment’, 49–118.

  35. 35 Note the successful rejection of the hypothesis in R.A. Carlson, David, the Chosen King: A Traditio-Historical Approach in the Second Book of Samuel (Uppsala: Almqvist and Wiksell, 1964) esp 131–39. A review of the problems of the thesis which in my judgment would gain even more force from the adoption of a Lakatosian analysis as a degenerating research programme is offered by Iain W. Provan, ‘On ‘Seeing’ the Trees While Missing the Forest: The Wisdom of Characters and Readers in 2 Samuel and 1 Kings', in Edward Ball (ed.), In Search of True Wisdom. Essays in Old Testament Interpretation in Honour of Ronald E. Clements (JSOTS 300; Sheffield: Sheffield Academic Press, 1999), 153–73, esp 156–62.