Show Summary Details

p. 926. So what is analytic philosophy?free

  • Michael Beaney


There are various similarities and differences between the respective approaches to analytic philosophy of Frege, Russell, Moore, Wittgenstein, and Stebbing. But is there anything in common that could be taken to characterize analytic philosophy as a whole? ‘So what is analytic philosophy?’ explains that analytic philosophy is ‘analytic’ in an extra special sense because it made use of modern logic together with all the new techniques that emerged in its wake and the greater understanding of the relationship between logic and language that this generated. It looks at later analytic philosophy—ordinary language philosophy, ideal language philosophy, and scientific philosophy—before considering what is wrong and good about analytic philosophy.

I said in the Introduction to this book that my aim was to take you on a thought-thinking trip rather than a sight-seeing tour. In the subsequent chapters we have explored five themes that illustrate analytic philosophizing as it was pursued by Frege, Russell, Moore, Wittgenstein, and Stebbing. We have asked certain questions and considered how the philosophers concerned answered them. As I hope has emerged, there are various similarities and differences between their respective views and approaches. But is there anything in common that could be taken to characterize analytic philosophy as a whole? I have argued that analytic philosophers prize clarity and precision. Would it not be a great irony if they couldn’t characterize their own discipline?

In my view, it is a mistake to try to define analytic philosophy in terms of any set of doctrines that are shared by all and only analytic philosophers. (Moore’s open question argument might well be applicable here.) Certainly, what counts as analytic philosophy today is so broad-ranging and open-ended that any attempt to do so would be futile. An alternative suggestion is to characterize analytic philosophy, more loosely, as held together by family resemblances, to use an idea from Wittgenstein’s later philosophy. This, too, is problematic, since between any two philosophers, further philosophers can always be found to construct a chain of family p. 93resemblances, so that every philosopher might end up counting as an analytic philosopher.

A more sophisticated answer would be to enrich the idea of a family resemblance by embedding it in a detailed historical story that explains the arguments, doctrines, and theories as they actually developed in the debates that were inspired by those recognized as the founders of analytic philosophy. Someone counts as an analytic philosopher to the extent that they actively and self-consciously contribute to these ongoing debates. I am very sympathetic to this approach, and have attempted to reflect it, in part, in this book—in discussing the way that Russell sought to solve the paradox he discovered in Frege’s work, for example, and the distinction between saying and showing that Wittgenstein came to draw in criticizing both Frege and Russell.

Telling such a historical story, however, is not incompatible with identifying certain features of analytic philosophy that can be helpful in understanding its main character and distinguishing it from other kinds of philosophy. What I have also tried to show in this book is that analytic philosophy is above all a way of doing philosophy, exemplifying certain virtues and using the new methods that came from the development of modern logic.

Analysing ‘analysis’

As I suggested in the Introduction, the obvious way to characterize analytic philosophy is in terms of its use of analysis. Given that ‘analysis’, in some form, has always played a role in philosophy, however, this just pushes the problem onto the question of what forms of analysis are employed in analytic philosophy. In Chapter 3, I distinguished ‘interpretive’ from ‘decompositional’ and ‘regressive’ analysis, and suggested that, while all three are present in analytic philosophy, it is the former that is especially characteristic of Frege’s and Russell’s work, and with which Moore was especially concerned in raising questions about analysis.

p. 94In Chapter 5, I mentioned Stebbing’s view of how metaphors can mislead us. Talk of ‘analysis’, in fact, provides an excellent example of this. One of the original meanings of the ancient Greek term ‘analusis’, from which ‘analysis’ derives, was ‘unravelling’: the prefix ‘ana’ meant ‘up’ and ‘lusis’ meant ‘loosening’ or ‘separating’. We find this meaning in Homer’s Odyssey, for example, where Penelope is reported as ‘unravelling’ by night the web of thread for a shroud p. 95she was weaving by day, in order to stave off her suitors, having promised to make a decision on whom to marry (in Odysseus’ long absence) when the web was complete (Figure 8). The term was then extended, metaphorically, in talking of ‘unravelling’ or ‘dissolving’ problems. (Penelope, after all, found a brilliant solution to her own problem!) In ancient Greek geometry, this soon acquired a more technical sense, referring to the process of working back to more basic theorems and principles, or simpler geometrical constructions, by means of which the relevant problem (proving a given theorem or constructing a certain figure) could then be solved. This is analysis in the ‘regressive’ sense, although exactly how it was meant to work has been disputed down the ages.

8. Penelope unravelling her web.

When Greek terms were translated into Latin, ‘analusis’ was rendered as ‘decompositio’, and both terms were transliterated into English as ‘analysis’ and ‘decomposition’. The two English terms were sometimes used as synonyms, sometimes with slightly different meanings—with ‘decomposition’ having more of the meaning of ‘breaking down’. (Something similar happened, most notably, with ‘fantasy’ and ‘imagination’, ‘ethics’ and ‘morality’, ‘psychological’ and ‘mental’, the first in each pair of terms deriving from the Greek, the second from the Latin. Sometimes they are used or understood more or less synonymously, sometimes as having rather different meanings. There is a similarly fascinating, philosophically significant, and long story in each case.) To the extent that ‘analysis’ and ‘decomposition’ were treated as synonyms, however, the connotations of ‘decomposition’ were projected back onto ‘analysis’, or perhaps more accurately put, reinforced the decompositional connotations that ‘analysis’ already had and obscured its other meanings. In other words, ‘analysis’ gradually took on more of the sense of ‘decomposition’, so that, today, the decompositional sense is indeed the one that is most associated with ‘analysis’.

However, the other senses lived on, and in particular, the key idea of analysis as ‘solution’ or ‘dissolution’ rather than ‘decomposition’ p. 96has also (to use yet another metaphor) infused conceptions of analysis ever since. Take the case of ‘analytic’ geometry, for example, which was created by Descartes and Fermat in the 17th century. What happens here is that geometrical problems are solved using the resources of algebra and arithmetic. A line, for example, is represented by the equation y=ax+b, where ‘x’ and ‘y’ give you the coordinates in plotting the line on a graph, with ‘a’ representing the gradient of the line and ‘b’ the point where the line cuts the y-axis (where x=0). Problems that cannot be solved within Euclidean (ancient Greek) geometry, which came to be called ‘synthetic’ geometry by way of contrast, can be readily solved within analytic geometry.

In analytic geometry, the geometrical problems are solved by ‘translating’ them into the language of arithmetic and algebra. And here we can also see how ‘interpretive’ analysis plays a role. Lines, circles, curves, and so on, must first be ‘interpreted’ as equations, and the geometrical problems correspondingly reformulated, before arithmetic and algebra can be applied in solving them. The idea here can be generalized: problems need to be interpreted in some form before the resources of a relevant theory or conceptual framework can be brought to bear. And this is exactly what is involved in analytic philosophy: the propositions to be analysed—those that give rise to the philosophical problems to be solved or dissolved—need to be rephrased in a richer conceptual framework or formalized in an appropriate logical theory. Analytic philosophy, then, is ‘analytic’ much more in the sense that analytic geometry is analytic than in any crude decompositional sense.

Let me stress again, though, that all three of the forms of analysis I have distinguished are typically involved in any actual example or project of analysis. Indeed, insofar as ‘decomposition’, ‘regression’ and ‘interpretation’ are themselves metaphors, they really just reflect different aspects or dimensions of the process of solving problems. We solve problems in all sorts of ways and combinations p. 97of ways. We may have to break them down into smaller parts that are easier to handle, to identify things we already know that may help us, and/or to reformulate them to open up new approaches. It is not surprising, then, that analysis is far more complicated than any particular metaphor is able to capture.

But if ‘analysis’, in its widest sense, just means problem-solving, then doesn’t all philosophy count as analytic? And if what we now call analytic philosophy is just ‘analytic’ in some extra special sense, then what is that sense that makes it appropriate to emphasize analysis in its self-description? The short answer here is the use it made of modern logic—the quantificational logic developed by Frege, Russell, and others—together with all the new techniques that emerged in its wake and the greater understanding of the relationship between logic and language that this generated. Just as analytic geometry transformed geometry by utilizing the powerful tools of arithmetic and algebra, so analytic philosophy transformed philosophy by utilizing the powerful tools of quantificational logic and theories of meaning.

Introducing later analytic philosophy

We have focused in this book on some of the main ideas of five of the founders of analytic philosophy. Thinking through these ideas gives a better sense of analytic philosophizing, in my view, than a whistle-stop tour of the whole terrain. But a lot has happened since the early years of analytic philosophy. So how has it evolved subsequently? Staying with the theme of analysis, let me outline two central strands in its later development by distinguishing two different ways in which the earlier conceptions and practices of analysis were transformed and extended.

In Chapter 5, we saw how a distinction came to be drawn—by Stebbing, among others—between logical or ‘same-level’ analysis and metaphysical or ‘new-level’ analysis. This distinction opened up the possibility of accepting the first while rejecting the p. 98latter—and especially, rejecting the assumption that there must be some final, definitive metaphysical analysis. This is the approach that was taken by what are now often called, though rather misleadingly, ‘ordinary language philosophers’ or ‘linguistic philosophers’, working especially in Oxford in the 1950s and 1960s, but including Wittgenstein, who rejected some of the key ideas of his Tractatus in developing his later philosophy in Cambridge in the 1930s and 1940s.

The main fruits of Wittgenstein’s later thinking are contained in his Philosophical Investigations, the main part of which was completed by 1945 but only published in 1953, after his death. Wittgenstein continues to be concerned with what distinguishes sense from nonsense, but he rejects his earlier assumption that there is a single logic—essentially that formulated by Frege and Russell—that underlies our use of language, as well as his earlier view that propositions have sense if and only if they picture possible states of affairs. Instead he emphasizes the multiplicity of what he calls our ‘language-games’, each of which is governed by its own ‘logic’ or ‘grammar’—the term he now prefers—understood as the set of rules according to which the language-game is played.

Philosophy, on Wittgenstein’s new conception, consists in gaining an ‘overview’ of the grammar of our language-games, aimed at clearing away the misunderstandings that may arise. In §90 of the Investigations, Wittgenstein writes:

Our examination is … a grammatical one. And this examination sheds light on our problem by clearing away misunderstandings. Misunderstandings concerning the use of words, arising, among other things, from certain analogies between the forms of expression in different areas of our language.—Some of them may be removed by replacing one form of expression by another; this can be called an “analysis” [“Analysieren”] of our forms of expression, for the process sometimes resembles decomposition.

p. 99As this suggests, Wittgenstein’s method might also be characterized as a form of ‘analysis’, although I think that he is being slightly misled himself in thinking of analysis too much in ‘decompositional’ terms. But it certainly counts as what Stebbing called ‘same-level’ analysis and I have called ‘interpretive’ analysis.

The second, rather different conception of analysis that developed after the Second World War was influenced more by Frege’s and Russell’s work than by Moore’s and Wittgenstein’s, and in particular, by the type of analysis exemplified in their logicist project. Recall Frege’s definitions of the natural numbers in terms of classes discussed in Chapter 1. These are not definitions that any ‘ordinary’ speaker is likely to come up with; as we have seen, they were intended to provide a deeper understanding of arithmetic by using the richer conceptual resources of the new logic. Carnap came to call this ‘explication’, which he characterized as involving the replacement of an ordinary, vague term by a more precise, scientifically defined term—in the way that we might replace the everyday concept of warmth by the more exact concept of temperature defined in physics, or our ordinary concept of water by the more precise concept of H2O defined in chemistry.

When the Nazis came to power in Germany in 1933, Carnap—together with other logical empiricists—emigrated to the United States, and the more ‘scientific’ conception of philosophy that they advocated took root there. Philosophy came to be seen as continuous with natural science, so that analysis, correspondingly, was viewed in a more scientific spirit, eventually opening the way—despite its repudiation in the heyday of logical empiricism—for reinvigorated metaphysical analysis.

After the Second World War, then, analytic philosophy can be seen as dividing into two main strands, one developing the approach of Moore and Stebbing, and the other building on the work of Frege and Russell, with Wittgenstein’s Tractatus having p. 100differing influences on both. Let us look at each strand in a little more detail.

Ordinary language philosophy

Besides Wittgenstein, three philosophers stand out as representatives of ‘ordinary language philosophy’: Gilbert Ryle (1900–76), J. L. Austin (1911–60), and P. F. Strawson (1919–2006), all of whom were based in Oxford. Ryle’s work epitomizes the move from new-level to same-level analysis in the development of analytic philosophy. In his ‘Systematically Misleading Expressions’ of 1932, he assumed that every proposition has a ‘correct’ logical form; but by the time he published his most influential book, The Concept of Mind, in 1949, he focused on charting what he called the ‘logical geography’ of our concepts. It was Ryle who popularized the idea of a ‘category mistake’, the central argument of his book being that the view, introduced by René Descartes in the 17th century, that the mind inhabits the body as a ‘Ghost in the Machine’, as Ryle described it, was a category mistake: what it makes sense to say about mental phenomena does not necessarily make sense to say about physical events, and vice versa. His book offered an account of a wide range of our concepts of mental phenomena, exploring their logical connections.

If Ryle was the Oxford Wittgensteinian, then Austin was the Oxford Moorean. His paper, ‘A Plea for Excuses’, published in 1956, offers the best introduction to his philosophical approach. He here addresses that perennial philosophical conundrum: do we have free will or is everything we do determined? Taking Moore’s line in seeking to distinguish the different questions that might be asked here, Austin suggests that the supposed opposition breaks down into a number of far more specific distinctions, which play out in much subtler ways in the concrete contexts in which the terms are applicable. Consider the range of adverbs we use in describing how someone has acted in a not entirely ‘free’ way, such as ‘inadvertently’, ‘mistakenly’, ‘accidentally’, ‘absent-mindedly’, p. 101‘unintentionally’, ‘carelessly’, ‘automatically’, ‘aimlessly’, and ‘purposelessly’. Is there really a single overarching distinction between acting freely and being determined? I shall (deliberately) leave you free to make up your own mind about this!

Austin laid the foundations for what we now know as speech act theory. He emphasized the different kinds of things that we do with words. In saying ‘I promise’, for example, I am not just saying something (picturing a possible state of affairs), but performing the act of promising: Austin called this a ‘performative’ utterance. Strawson took up this idea of a speech act in criticizing Russell’s theory of descriptions (as first presented in ‘On Denoting’ of 1905) in an influential article he published in 1950 entitled ‘On Referring’. According to Strawson, if there is no King of France, then the statement that the King of France is a philosopher should be regarded as neither true nor false, not false as Russell had claimed. I am not asserting that there is one and only one King of France, but presupposing it as a condition for asserting something about this person. Assertion and presupposition are speech acts, something that we do in using language, and Strawson’s basic charge against Russell is that he focused on sentences and not on our uses of sentences.

Strawson’s later work is characterized by a return to metaphysics, but one that he calls ‘descriptive’ rather than ‘revisionary’, aimed at clarifying the fundamental conceptual framework by means of which we think about the world. In Analysis and Metaphysics, published in 1992, he distinguishes between ‘connective’ and ‘reductive’ analysis and endorses the former: the philosophical elucidation of a concept consists in explaining its often complex connections to other concepts.

Ideal language philosophy and scientific philosophy

Let me single out three philosophers to give a sense of developments in analytic philosophy in post-war America: W. V. O. Quine p. 102(1908–2000), Donald Davidson (1917–2003), and Hilary Putnam (1926–2016). Quine’s most influential paper is ‘Two Dogmas of Empiricism’, published in 1951, in which he attacked the logical empiricists’ distinction between analytic and synthetic propositions as well as their reductive conception of analysis, according to which synthetic propositions are ultimately verified by sensory experience. According to Quine, there is no clear distinction between matters of ‘meaning’ and matters of ‘fact’, and he proposed an alternative view on which our whole ‘web of beliefs’, as he called it, is empirically tested as a whole.

Quine agreed with Carnap, however, that the role of philosophy was to ‘explicate’ our ordinary concepts and beliefs by ‘translating’ or—as he put it—‘regimenting’ them into a suitable logical language. But while Carnap had seen this as a way of eliminating metaphysics, Quine regarded this—as Russell had done earlier—as revealing our ‘real’ ontological commitments, in other words, as showing what our best scientific account tells us about what things there are in the world. In Strawson’s terms, Quine endorses ‘revisionary’ metaphysics, albeit informed by the results of the natural sciences.

Davidson was highly influential in another major development in post-war analytic philosophy: the construction of theories of meaning. Frege’s work is generally taken as the starting-point, but Frege’s concern was primarily with mathematics—with how mathematical language expresses the thoughts and arguments it does. Davidson was concerned with constructing a theory of meaning for natural languages, in all their richness and complexity. This is now a huge research industry, as both philosophers and linguists seek to understand how the various elements of language work, from pronouns to metaphors. One contribution Davidson made was to the analysis of sentences about actions and events, which he suggested showed how we are committed to an ontology of events—in other words, that events p. 103should be taken as existing in the world just as Frege, for example, thought that objects and concepts exist.

Davidson also identified and rejected what he called the third dogma of empiricism: the view that one can distinguish between conceptual scheme and empirical content. According to Davidson, we can make no sense of someone’s having a completely different conceptual scheme from our own. If true, then this has implications for the question we considered in Chapter 3: we can only know what another person means if we have some common conceptual understanding, rooted in our shared engagements with the world. Grasp of meaning, the ability to interpret others, and participation in collective activities are fundamentally interconnected.

Putnam has had a significant influence on American philosophy across a wide range of fields. I shall just mention here his thought experiment about Twin Earth, since this has relevance for the issue we considered in Chapter 3 concerning the ‘analysis’ of water as H2O. Twin Earth is imagined as exactly like our Earth except that what looks and tastes like water is actually not H2O but a different chemical compound, say, XYZ. According to Putnam, and the ‘intuitions’ about what to say here that he hopes to invoke, someone on Earth would be referring to and thinking about H2O, and someone on Twin Earth would be referring to and thinking about XYZ, even if neither of them knew what the chemical compound was of the watery substance they were experiencing. So meanings and thoughts are partly individuated by what lies outside us, in the external world, a thesis that is called ‘externalism’. If this is right, then there is a response to the paradox of analysis: the ‘external’ meaning helps secure the correctness of ‘Water is H2O’, and yet you can still learn something when you discover this. Putnam captured his basic idea with the slogan ‘meanings just ain’t in the head’. The debate about externalism has raged ever since, both inside and outside people’s heads.

How did analytic philosophy get its name?

We now see the origins of analytic philosophy as lying in the work of Frege, Russell, Moore, and Wittgenstein in the four decades around the turn of the 20th century. (Frege’s Begriffsschrift was published in 1879, Wittgenstein’s Tractatus in 1921.) Yet we do not find the term ‘analytic philosophy’ being used to refer to even part of what we now recognize as the analytic tradition until the early 1930s. The term does not appear, for example, in what I suggested is the first textbook of analytic philosophy—Stebbing’s A Modern Introduction to Logic, published in 1930. We should not find this surprising, however, since it takes a certain amount of time for any movement or tradition to establish itself sufficiently to be conceptualized as such and for any name to catch on.

What may be surprising, though, is that the term is first used in criticizing analytic philosophy. The Oxford philosopher R. G. Collingwood (1889–1943) uses it in his Essay on Philosophical Method of 1933, in attacking the view according to which philosophy seeks only to analyse what we already know, a view which is attributed to Moore and Stebbing. For Collingwood, ‘analytic philosophy’ was intended as a derogatory term, suggesting a narrow-minded conception of philosophy. As we saw in discussing the paradox of analysis in Chapter 3, the question of how analyses can be informative is indeed a fundamental issue, so there is something to Collingwood’s criticism. However, as I hope I showed, appreciating the transformative character of analyses offers an effective response to the paradox (a response, in fact, that we can find in Collingwood’s own Essay).

The group of philosophers to which Moore and Stebbing belonged did indeed have a name in the 1930s—the Cambridge School of Analysis. The name reflected the inspiration that Russell’s programme of analysis and Wittgenstein’s views on analysis had on their work. So it was entirely appropriate to use the term p. 105‘analytic philosophy’ as an alternative name, understood more positively. In the second half of the 1930s the term was extended to include the logical empiricists, who had also been influenced by Russell and Wittgenstein; but it was only after the Second World War that the term really caught on. By then it was also used to include both ordinary language philosophy and the new forms of ideal language philosophy and scientific philosophy developing in the States.

During the 1950s both the influence that Frege had on Russell, Wittgenstein, and Carnap, and the importance of his ideas for the project of constructing theories of meaning began to be appreciated, and analytic philosophy was ‘backdated’ to include him. Since then the canon has been expanding more and more, both backwards, sidewards, and forwards, and related philosophical traditions such as American pragmatism have been co-opted as well. (No one has yet been fired from the canon!) What began as a set of ideas and methods rooted in logic, philosophy of mathematics, and ethics, has now expanded through philosophy of language, philosophy of science, and philosophy of mind to all areas of philosophy. For every branch of philosophy, there is now an ‘analytic’ version, not just analytic metaphysics and analytic aesthetics (dating from the 1950s) but analytic Marxism, analytic phenomenology, and analytic feminism, for example, and I have even seen the term ‘analytic dogmatic theology’ being used! The name ‘analytic’ has caught on to such a degree that one might wonder what does not now count as analytic philosophy, at least in Western philosophy. To what is it opposed?

Analytic and continental philosophy

In 1958, a conference was held at Royaumont in France to which various analytic philosophers were invited to encourage dialogue with French philosophers. The conference was not a great success. Gilbert Ryle read a paper in which he talked of the wide gulf that p. 106had opened up between Anglo-Saxon and Continental philosophy. By ‘Anglo-Saxon’ philosophy he meant analytic philosophy and by ‘Continental’ philosophy he meant phenomenology, in particular. Since then the meaning of ‘Continental philosophy’ has broadened to include other traditions such as hermeneutics, existentialism, deconstruction, and indeed all kinds of Western philosophy from Kant onwards that are not analytic. The division has become one of the most entrenched and pernicious in contemporary philosophy. It illustrates, though, how traditions (like political parties) are partly formed and maintain themselves in mutual opposition.

As Bernard Williams once famously remarked, distinguishing between ‘analytic’ and ‘continental’ philosophy is like dividing cars into front-wheel-drive and Japanese; one term is methodological or operational, the other geographical. The supposed contrast involves a kind of category confusion, and is neither exclusive nor exhaustive. Recognized analytic philosophers such as Frege, Wittgenstein, and Carnap, for example, were brought up—and in Frege’s case, spent his whole life—in German-speaking countries. And ‘continental’ philosophy employs forms of analysis as well. There are also many philosophical traditions, such as pragmatism, not to mention all the various non-Western traditions, that are hard to classify as either. However, just as we can unpack the metaphor of ‘analytic’, so too we can unpack the metonym of ‘continental’. ‘Continental philosophy’ is an umbrella term for a range of traditions that have their origins—unlike analytic philosophy—primarily in the work of philosophers based in continental Europe. The founder of phenomenology was Edmund Husserl (1859–1938), for example, and key figures in the development of hermeneutics include Friedrich Schleiermacher (1768–1834) and Hans-Georg Gadamer (1900–2002). Existentialism has been most associated with the work of Jean-Paul Sartre (1905–80) and Simone de Beauvoir (1908–86), but has roots further back in the work of Søren Kierkegaard (1813–55), Friedrich Nietzsche (1844–1900), and Martin p. 107Heidegger (1889–1976). Jacques Derrida (1930–2004) has been the guiding spirit of deconstruction.

9. A philosopher driving a front-wheel-drive Japanese car.

All the traditions that make up ‘continental philosophy’ have little in common beyond, negatively, their (supposed) opposition to analytic philosophy, and, positively, their all being, in some sense, responses to Kant’s philosophy (though this would also include analytic philosophy). Since this book is about analytic philosophy, I will say no more about continental philosophy directly but instead consider some of the criticisms that continental philosophers, among others, have made of analytic philosophy, which have helped maintain the supposed opposition.

What is wrong with analytic philosophy?

The criticisms that continental philosophers have made of analytic philosophy, as well as those that analytic philosophers have made of continental philosophy, have frequently involved caricatures of the views they oppose. In Box 6 I have listed some of the main differences that have been claimed to hold between analytic and continental philosophy, all of which are caricatures, although—like all caricatures—they contain some element of truth that makes them plausible. Some concern method and style, with which we have been concerned throughout this book and to which I will return in the final section. I shall say something about the p. 109other differences here, under the two broad headings of the relationship of analytic philosophy, first, to natural science, and second, to history.

Box 6. Some caricatures of the analytic/continental divide

Analytic philosophy

Continental philosophy

Uses analytic methods

Uses synthetic methods

Clear, precise, rigorous

Suggestive, allusive, playful

Logical argumentation

Sceptical of powers of reason



Concern with truth

Concern with meaning















Especially over the last three or four decades analytic philosophy has been seen and criticized as both scientistic and naturalist. Scientism is the view that the natural sciences provide the model for explanation and understanding in other fields, including philosophy. Naturalism comes in two main varieties. In its stronger form it is similar to scientism, holding that everything can ultimately be explained by natural science. In its weaker form, it holds that there is nothing over and above the natural world—in other words, it rejects all appeals to anything ‘supernatural’.

From Russell and the logical empiricists onwards, there has been a strong tradition within analytic philosophy that can certainly be described as ‘scientific’, in its respect for the methods and results of the natural sciences and in its weak naturalism. But that is not the same as being scientistic and strongly naturalist, though some analytic philosophers are that as well. It depends on one’s view of the place of logic and metaphysics in all this. Many philosophers of mind today draw on and engage with the latest work in biology, cognitive science, neurology, and psychology, and many philosophers of language are in continual dialogue with linguists, among others. But they do so in a critical spirit and will use other methods than those of the natural sciences, so ‘scientistic’ is too strong a term to use even for them.

There is also a different strand that goes back to Frege, Moore, and the early Wittgenstein, who all rejected scientism and naturalism, although the later Wittgenstein is a naturalist in the weaker form. Frege conceived numbers as logical objects, understood as non-natural objects, for example, and Moore was a non-naturalist about ethics. The debate between naturalism and non-naturalism is a major one in all areas of contemporary p. 110analytic philosophy, so it would be wrong to see either position as characteristic of analytic philosophy.

In my view, there is one fundamental objection to scientism (and strong naturalism), which can be stated like this. Any science makes presuppositions, and it has traditionally been one of the primary tasks of philosophy to subject these to critical examination, tasks that require different methods to those of science itself. Analytic philosophy is well placed to do this, but not uniquely so. Phenomenologists, in particular, have offered a powerful critique of scientism, in arguing that the whole project of natural science, and the theoretical attitude it requires, needs to be understood as emerging from what Husserl called our ‘life-world’ (‘Lebenswelt’), comprised of our everyday human practices and pre-theoretical attitudes and beliefs. In my view, analytic philosophy has not fully learnt the lessons of phenomenology, so this is one area where criticisms of analytic philosophy are justified.

The other main, though related area in which I would say that criticisms of analytic philosophy are justified concerns its attitude to history. Analytic philosophy has often been accused of being ‘ahistoricist’ and even ‘anti-historicist’. Like naturalism, historicism comes in various forms, the strongest of which claims that philosophy is essentially historical and needs to be understood and pursued as such, and the weakest of which simply claims that knowledge of history is useful in pursuing philosophy. Few dispute the latter, but the former is far more controversial. Many analytic philosophers, and certainly the early ones, did indeed reject historicism in its strongest forms. Philosophical problems, it was assumed, are timeless, and while it might be useful to see how previous philosophers sought to solve them, fresh attempts can always be made, such as by drawing on the latest logical or scientific theories.

Let us grant that there are some philosophical problems that are universal, if not exactly ‘timeless’—because they are rooted in such p. 111fundamental human activities as counting and inferring. We can also agree that solutions to these problems may involve the fashioning of new conceptual tools. Nevertheless, historical understanding will still be required. For the new concepts and solutions will need to be explained and this will involve making clear how they are related to other concepts and solutions, both past and present. As philosophical debates proceed, engagement with previous views is inevitable.

Philosophical views, too, make presuppositions, as we saw in Chapter 5. These may not be clear at the time the views are first formulated or debated; but sooner or later they will need to be made explicit, if the debates are to progress. It may require a certain historical distance to identify these presuppositions, so here again is an area where historical understanding is needed.

Historical understanding may also be required in unpacking the metaphors, analogies, parables, allegories, and similes that also play a role in philosophy, as I hope I have illustrated in clarifying the idea of analysis itself; and to these we might add all the allusions and references—for example, to the ideas of past thinkers—that we also find in philosophical writing. Derrida (one of the continental philosophers mentioned above) has drawn our attention to what he calls the ‘margins’ of philosophy—to footnotes, prefaces, letters, interviews, casual remarks, and so on—where a lot of ‘unofficial’ thinking takes place, which may shed light on what is ‘officially’ said. Here again is where the work of the historian comes into play.

Analytic philosophers often criticize historians of philosophy for ‘merely’ being concerned with what a past philosopher meant or thought, focusing on interpreting texts, while their concern, they say, is with what is true, focusing on solving philosophical problems. But these two concerns are not as distinct as they suppose. If we are to interpret what someone means as charitably as possible, then we will need to know what is true, and any evaluation of their p. 112views will certainly require this. And solving problems and finding out what is true may be fostered by working through the solutions and views that others have offered, which will require understanding what they meant. Analytic philosophy and history of philosophy need each other.

Lack of historical self-consciousness is arguably the biggest blind-spot in analytic philosophy. Not being overawed by tradition may be essential in making conceptual innovations, but establishing and defending those innovations requires locating them in the historical space of previous views. Continental philosophy offers rich resources for understanding this historical space—and for uncovering presuppositions, unpacking metaphors, deconstructing assumptions, contextualizing attitudes, and so on. Here is where analytic philosophy would benefit most, in my opinion, from deeper engagement with continental traditions.

So what is good about analytic philosophy?

On the account that I have offered in this book, analytic philosophy has its origins in the development and use of modern logic by Frege, Russell, and others. Philosophical problems are solved by translating the problematic sentences into logical language, making clearer what they ‘really’ mean. I described this process as involving interpretive analysis, as well as the regressive and decompositional forms of analysis that have always been present in philosophy. Interpretive analysis has been fundamental to analytic philosophy ever since, opening up questions about the nature of meaning and the relationship between logic and language that have been central themes throughout its history. So the first thing I would say is good about analytic philosophy is the deeper understanding it has fostered of the complex and myriad ways in which language works.

The development of modern logic has changed the philosophical landscape irrevocably. Before Frege, logicians had only been able p. 113to analyse a small subset of the inferences we make in everyday and scientific thinking. Building on Frege’s work, logical and semantic theory has been further developed to deepen our understanding of the whole range of our conceptual and reasoning practices. It has also enabled philosophers to set out arguments as rigorously as possible, identifying all the premises and rules of inference that allow the conclusion to be validly drawn. By reconstructing arguments offered in the past, philosophers can demonstrate their validity or else show where they fail or what missing premises are needed. Applying logic—and the ideas of analytic philosophy, more generally—has deepened our understanding of the history of philosophy, mathematics, and science, both Western and non-Western.

Let us return, though, to the virtues which I mentioned at the beginning of this book, and which have been themes in the subsequent chapters. I hope that I have shown how clarity of thinking, precision of expression, and rigour of argumentation came to be regarded as central virtues in analytic philosophy, and justified my own belief that they are indeed virtues. I have also sought to illustrate how analytic philosophy is conceptually creative, something that deserves more recognition. When one looks back at the history of analytic philosophy, one realizes just how intellectually fruitful its conceptual innovations have been. Systematic work has also been done, from Frege’s and Russell’s logicist project to contemporary theories of meaning. But analytic philosophy also lends itself to more piecemeal contributions, such as in analysing specific concepts or reconstructing or criticizing particular arguments. A wide range of conceptual and logical tools have now been developed to assist the analytic philosopher’s work, and what is first used in one domain has been applied in other fields.

In recent years, analytic philosophers, at least in some quarters, have become more historically self-conscious and increasingly willing to engage in dialogue with other philosophical traditions, p. 114from ancient Chinese philosophy to French deconstruction, and these developments are to be welcomed. As I have indicated, analytic philosophy has also ramified into all areas of philosophy, from logic and philosophy of mathematics to theology and critical thinking, and there is exciting work being done, drawing on, refining, and enlarging its methodological toolbox. The future of analytic philosophy is clear, even if its precise nature and contours are a matter of argument.