‘Believer’ is a synonym for an adherent of a religion in ordinary English, and scholars have for some time worried about the Christian roots of this association. But their concern is often rather inchoate. Christianity is an enormously heterogeneous phenomenon. Is the association between ‘belief’ and ‘religion’ simply Christian, or should it be identified with particular Christian traditions, or with particular periods in Christian history? It is also not clear what follows if the association between belief and religion is distinctively Christian, especially for the use of the concept in the study of non-Christian religions. Christians have often foregrounded belief when judging whether someone counts as a genuine Christian. Yet it does not follow that non-Christians do not have beliefs, or that what people believe does not – to some degree at least – define them as the beings they are.
In addition to these specific problems we might ask a broader question about the historicity of belief. We are familiar with claims that particular psychological states are culturally specific and therefore historically contingent. Could this be true of belief? Several scholars have hinted at the possibility. Thus in his essay ‘Christians as believers,’ Malcolm Ruel accepted Rodney Needham's contention that belief is not a universal human experience, but suggested against Needham, who held that belief must be universal or nothing at all, that there was such an experience for Martin Luther (Ruel 1982: 25; Needham 1972: 151, 234). More recently a number of anthropologists and scholars of religion have argued that the ascription of beliefs to pre-Christian Europeans – even sometimes pre-modern Christians – or to people from areas ‘outside the historical sphere of Christian dominance’ is ethnocentric and/or anachronistic, either because ‘the concept derives from modern Christian usage,’1 or because it has roots in modern secularist thought.2
These claims often appear alongside (and not always clearly distinguished from) other arguments that have more modest implications – for instance, that the emphasis we place on belief when thinking about religious life is unusual and culturally specific, an argument compatible with a universalist conception of belief.3 It is therefore unclear whether scholars positing the historical and cultural contingency of belief are committed to the implications of their claims, or indeed whether they know what those implications are. But the claims are there, and they are, potentially, important enough to warrant careful assessment.
My concern in this article is to determine whether a true and informative case can be made for them. I emphasize informative because the claim that belief ‘has a history’ can be made true by definitional fiat, but it will only be informative if it is about belief as ordinarily understood. For ‘belief’ and ‘believe’ are not technical terms that anthropologists and historians learn to use in the course of their training, and they are not ordinary terms that have technical uses within these disciplines in the way that, say, terms like ‘force’ and ‘mass’ have technical uses within physics. There is therefore no anthropological or historical concept of belief. When Peter Brown tells us that the fourth century Roman aristocrat ‘Symmachus sincerely believed that invisible beings hovered above Rome’ there is no reason to think that ‘believe’ means something different here – that it is ascribed in virtue of different criteria – from what it means in such statements as ‘70–80 percent of consultants and similar percentages of junior doctors believe that safe patient care is undeliverable in a 48-hour week,’ or ‘The prime minister believes that provoking chaos will get him a better deal in EU negotiations’ (Brown 2012: 105).4
Much therefore hangs on the question of how the terms ‘belief’ and ‘believe’ are ordinarily used and understood, and on how we are to achieve an accurate reflective conception of our ordinary use and understanding of these words. Over the course of this article I will suggest that scholars proposing the more ambitious arguments for the historical and cultural contingency of belief have been insufficiently attentive to this question, and that their arguments rest on misrepresentations of the meaning of belief ascriptions. As such they have not given us any compelling examples of cases where the ascription of beliefs to historical and/or cultural others is anachronistic or ethnocentric, and it is difficult to see how an argument of this kind could be made.
Having a History
Claims of the form ‘x has a history’ often promise more than they deliver. Thus it would be surprising and disquieting to learn that truth has a social history, as Steven Shapin has implied. But Shapin's concern is really with the social history of what people have taken to be true, a much less philosophically contentious subject (Shapin 1994: 3–5). A similar overselling attaches to Ethan Shagan's recent book on The Birth of Modern Belief, which by his own admission aims to do for religion what Shapin did for truth (Shagan 2018: ix).
Shagan begins by asserting that belief ‘has a history,’ and he emphasises that his concern is with ‘the history of belief itself’ (Shagan 2018: ix, 100). But he quickly shifts from ‘the history of belief itself’ to the history of the ‘category’ or ‘concept’ of belief (Shagan 2018: 10–11, 18–19, 26). These expressions are not synonymous, and it would take a good argument to show that a claim about the historicity of the category or concept of belief has implications for the historicity of belief itself.5 But Shagan makes no such argument. Indeed, he nowhere establishes what he understands the concept to be, so it is not clear what his book is about.6 My own view is that it is a history of ideas about belief, particularly religious belief, and Shagan explores this subject to considerable effect. But he does not show that belief itself has a history, and I am not sure he shows that the concept of belief has a history either, at least not in the sense that its content has changed over time.7
I start with Shagan because his book inadvertently highlights the need to be clear about what we mean when we say that belief ‘has a history.’ For the expression is open to several interpretations, some banal (what people have thought about belief ‘has a history’), and some more radical, even bewildering (‘the concept of belief has a history,’ or ‘belief itself has a history’). I want to start, however, by looking at some general features of ‘x has a history’ claims, their use and possible meanings.
The purpose of asserting that something ‘has a history’ is to ‘raise consciousness’ (cf. Hacking 1999: 5–6). No one needs to tell us that Christianity ‘has a history,’ because it is common knowledge. When someone asserts that x has a history, they want to unsettle our attitude toward x and ultimately to change our understanding of x, by showing that x was once importantly different, or that it once did not exist, the claim's effectiveness depending on our regarding x as unchanging or universal, or our having no critical attitude toward x and taking its existence for granted.
With respect to meaning, ‘x has a history’ is ambiguous between two claims: first that x has changed over time, and second that x emerged at some point in time. These claims are not mutually exclusive – Christianity has a history in both senses – but they have different implications. For the first is equivocal with respect to x's historical identity, since it can only be sustained if the historian can identify x at different times, implying that x has core features that are, if not necessarily invariable, then stable enough to sustain identification and reidentification. When historians argue that particular emotions ‘have a history’ they are often making this sort of claim. They are not saying that, for instance, people in antiquity did not feel anger, but rather that ancient anger was importantly different from contemporary anger: it had a different intensity, was prompted by different sorts of events, had different behavioural manifestations, and so on. But it must share enough in common with anger as we understand it for us to be able to call it anger, and not another emotion altogether.8
When scholars assert that x ‘has a history’ they imply that hegemonic or untutored conceptions of x are mistaken. Shagan's assertion that ‘belief has a history: it changes over time’ is not a truism (Shagan 2018: 1). But it is also not obvious how we should assess its truth, since the identity of the ‘it’ – belief – that is being said to change over time is not clear. It is enough to note that Shagan's claim can only be vindicated if we can identify belief in the middle ages, early modern Europe, and modern Europe, and what we identify must be belief as we understand it. The question ‘what is belief?’ is therefore prior to any historical question, and if it is not addressed explicitly an answer will be presupposed.
The assertion that something has a history in the sense that it emerged at some point in time implies a more robust historicism, and assertions of this kind can be startling. Consider Elizabeth Goodstein's argument that boredom is an experience ‘peculiar to modernity,’ which implies that people in antiquity, the middle ages, and early modern Europe did not feel boredom (Goodstein 2005: 3–7, 18, 46; 2020: 26). But claims of this kind also face a ‘what is x?’ question, which again is prior to any questions about x's historicity, for we must know what x is if we are to chart its emergence.9 Goodstein's claim can only be informative – and genuinely surprising – if it is about boredom as we ordinarily understand it, boredom being a term of ordinary language. And this raises the question of how we are to determine what we understand boredom to be. It seems to me that a credible answer to this question will have to be consistent with a plausible account of how we learned to use the word, and it must also be able to accommodate the mundane contexts in which we ordinarily speak of being bored. I doubt that Goodstein's characterization of the feeling can do this, and I suspect therefore that her startling claim is largely secured by her definition of boredom.10
The claim that belief ‘has a history’ in this more radical sense is a startling one, which implies that whole programs of anthropological, psychological, and historical research are based on essentially parochial premises. But proponents of this claim face a daunting burden of proof, for they would have to show scholars in these fields that their use of the word carries commitments that they are unaware of. And these commitments must be more than connotations, for it is not surprising to learn that ‘belief’ has culturally specific associations. Jason Davies may be right that ‘when we talk about “religion” and “belief” we are generally drawing on a predominantly Christianised perspective that emphasised inner experience, spirituality, the well-being of the soul (as more important than the body) and some kind of “core belief” (which is suspiciously similar structurally to the Catholic creed)’ (Davies 2011: 397). But what he describes is not anything that talk of ‘religion’ and ‘belief’ can plausibly be thought to entail. Indeed, Davies himself wants to say that there is something we can call ‘Roman religion’ that is not marked by concerns with ‘inner experience, spirituality, the well-being of the soul . . . and some kind of “core belief”.’
To vindicate the argument that belief has a history in the sense that it could be anachronistic to say that people somewhere believe, we would have to show that the use of the term entails a set of historically contingent commitments. It is not obvious how this could be done, but I suspect that the commitments would not be located at the level of the word's lexical meaning, for it is difficult to see how people can be systematically ignorant about the meaning of their own words (as distinct from their connotations), but rather at the level of the conditions that sustain the word's use.
Histories of Belief
My interest is in the viability of a robust historicism about belief, but I want first to discuss three more modest senses in which belief may be said to have a history, for I suspect the more radical claims for its historicity extrapolate from these. They are:
Belief has a history in the sense that what people believe changes over time;
Belief has a history in the sense that what it is to believe something can change over time;
Belief has a history in the sense that thinking about belief has changed over time, in ways that may have implications for certain uses of the term ‘believe.’
The first sense in which belief ‘has a history’ is mostly uncontroversial, although its scope will be affected if belief is historically contingent in a stronger sense. For if belief is a modern phenomenon, we would not want to say that the Romans had different beliefs from us: the claim would be that they did not believe anything. But for the most part the notion that beliefs vary across time and culture is a commonplace within the humanities and social sciences, with debate surrounding the extent of the variation.
The second sense in which belief may be said to have a history concerns particular beliefs or classes of belief, and can be developed in several ways. One is by focusing on changes in the practical, intellectual, and institutional context in which the beliefs are held. So, for example, in his classic study of Rabelais, Lucien Febvre argued that the ‘mode’ (façon) of theistic belief in the sixteenth century was fundamentally different from now, because everyday life then was structured by Christian practice and belief (Febvre (1942) 1982: 5, 336). Febvre wanted particularly to show that ‘unbelief’ in the contemporary sense was all but unthinkable for Rabelais and his contemporaries, because there was no way to justify the position: the putative sixteenth century atheist ‘had no ground on which to stand,’ and his denials of the existence of God ‘could at best have been no more than opinions – paradoxical ways of thinking and feeling that nothing from outside came to the support of or propped up in any real or substantial way, nothing in either the science or philosophy of his time’ (Febvre (1942) 1982: 461). More recently Charles Taylor defended a similar position, arguing that ‘belief in God isn't quite the same thing in 1500 and 2000’ owing to changes in the ‘conditions’ of that belief, or the ‘context or framework of the taken-for-granted’ in which it is held (Taylor 2007: 13). People in 1500 lived in a world that ‘told in favour of belief, made the presence of God seemingly undeniable’ (Taylor 2007: 25). Now, however, belief in God is open to reasonable challenge and must be held in light of this fact.
Febvre and Taylor may have overstated the all-encompassing character of Christian belief in the fifteenth and sixteenth centuries (Sheehan 2010: 222–25), but it seems hard to doubt that there are important differences between what it was to be a theist in early modern Europe and what it is to be a theist now. One way to think of these differences is by reference to the historical variability of belief in sense (i), for Taylor's ‘conditions of belief’ include other beliefs (about the natural world, political authority and so on), while a large part of Febvre's argument for the impossibility of atheism in the sixteenth century is that people then lacked the intellectual resources with which to formulate and justify the position. Or to put the matter more generally, there are radical differences between the beliefs of the average contemporary European or American and the average European of the sixteenth century, and this fact has implications for the significance of the beliefs they share, for these will be situated within different contexts of belief. Accordingly the way in which the shared beliefs will be justified, the inferences that the believer can draw from them, and the role the beliefs play in guiding action may differ significantly between historical periods.
Another way to think about the historical variability in what it is to believe something is suggested by certain pragmatist theories of classification.11 William James is the progenitor of this line of thought, and his views on classification start from the claim that we ‘can neither think nor experience’ reality as such because it ‘exists as a plenum,’ whose ‘parts are contemporaneous, each . . . as real as any other, and each as essential for making the whole just what it is and nothing else’ (James (1890) 1981: 1231; Leary 2018: 264). To make sense of things we must first break reality up into various ‘orders’ (James also speaks of ‘worlds’ and ‘sub-universes’), the orders being determined by our interests and purposes. It is within these orders that we form and hold our beliefs. We believe that Ivanhoe married Rowena within the ‘Ivanhoe-world.’ We believe that the table is a metre from the armchair within what James calls the world of sense, but not within the world of physics, where there are no tables and chairs, only particles and fields of force (James (1890) 1981: 917–27).
James was a pluralist about ‘orders of reality’ – there are, he thought, as many possible orders of reality as there are possible human interests and purposes – and none of the worlds I have mentioned reduces to the others, at least not epistemically. We can only form beliefs about Ivanhoe and Rowena, tables and chairs, and particles and fields of force, in particular ways, each of which presupposes a particular set of ontological commitments and methods of inquiry. This is, I suppose, trivially obvious in the first instance, the only way to find out whether Ivanhoe married Rowena being to read Ivanhoe. But the point has more interest in the second example, for James's position implies that the physicist cannot show that the table is not really one metre from the armchair, because qua physicist, he cannot form beliefs about tables and armchairs.
James's concern was with the heterogeneity of our beliefs at any given moment, but his ideas can be given a historicist turn, by registering changes in the order of reality within which a belief is held. To return to Febvre's discussion, the distinctions we might draw between religious and non-religious beliefs will have been drawn differently in the sixteenth century, if they were drawn at all. Consider only the fact that, among Rabelais’ contemporaries, the Bible was standardly understood to be a truthful historical record (among other things), and furthermore a record of incomparably greater epistemic authority than any secular text. Noah, Abraham, and Moses were therefore real historical individuals, no different in their mode of being from Julius Caesar or Augustus. The Bible no longer has this status for many Christians, and whatever beliefs they have about the life and deeds of Moses will have a different character from their beliefs about the first Roman emperors.
The two senses in which belief can be said to ‘have a history’ that I have explored thus far offer no support for the idea that belief itself is historically contingent, since they imply that people everywhere have beliefs. The third sense in which belief has a history is associated with some of the concerns that have arisen about the concept's use in cross-cultural interpretation. We may start with the uncontroversial observation that the Christian tradition has attached an unusual importance to belief, especially to the individual's having the correct beliefs about God, Christ, the virgin Mary, the nature of evil, and so on.12 The notions of orthodoxy and heresy were formulated during the second century CE, and they structured disputes between Christians for at least the next millennium and a half.13 Christians have therefore – to varying degrees – defined themselves by reference to their beliefs in a way that had no precedent of comparable sociological significance in pre-Christian antiquity. Indeed, it would seem odd to call Romans of the late republic ‘believers’: not because (or at least not necessarily because) they did not have beliefs, but because their relationship toward their beliefs differed from the relationship suggested by the expression ‘believer’. Above all, there was no sense that being Roman involved adhering to certain beliefs.
Now it is not clear that we can draw any conclusions about the interpretive use of the concept of belief from its unusual significance within the Christian tradition. William Robertson Smith long ago argued that ‘when we approach some strange or antique religion’ we should not assume that ‘our first business is to search for a creed, and find in it the key to ritual and practice’ (W.R. Smith 1894: 16–17). But he rightly did not take an absence of creeds to entail an absence of beliefs, and spoke of ancient Semitic beliefs about the kinship between gods and men or about the jinn (W.R. Smith 1894: 43–46, 119–39). It is therefore not obvious why students of non-Christian religions should not attend to questions of belief even if there is no creed or discourse about beliefs within those religions. At the same time students of historical forms of Christianity may find that the Christians they study did not share many core theological beliefs, and that their life as Christians was held together more by shared practices and institutions than by shared beliefs (indeed, Christian efforts at reform have not tended to reveal a deep consensus among Christians about their core intellectual commitments). It is therefore not obvious what importance a scholar of Christianity should attach to beliefs, even if Christians themselves emphasise their importance.14
Some of the confusion that surrounds these questions about the importance of belief stems, I think, from a slippage that can occur between two claims: first that Christians (or at least, Christian ideologues) attached an unusual significance to belief, and second that belief itself is a Christian concept. Consider Ruel's essay “Christians as believers.” When Ruel talks of the ‘monumental peculiarity of Christian “belief”,’ what does he mean by ‘Christian belief’ (Ruel 1982: 9)? The peculiarity of Christian ideas about belief – about its importance, say – or uses of the term ‘belief’? Or is he claiming that belief itself is peculiarly Christian?
While Ruel tends toward the first position – which is the position his history supports – he never settles on it, and there are hints throughout his essay of the second and more radical argument. He begins with a quote from Wilfred Cantwell Smith, which raises the problem of how western students of non-Christian religious groups should deal with ‘the peculiarity of the place given to belief in Christian history,’ and he criticizes ‘unsuspecting Western anthropologists who . . . give primacy to what people “believe” without fully declaring what the word means, nor recognizing, it would seem, just how rooted the concept is in our own cultural religious tradition, Christian and post-Christian, and thus how loaded any statement concerning “belief” easily becomes’ (Ruel 1982: 9; my emphases). These formulations imply that ‘belief’ has Christian connotations, or perhaps that there are distinctively Christian uses of the word, but not that it is implicitly Christian at the level of its meaning. Yet Ruel's talk of the concept as ‘rooted in’ our own cultural religious tradition could also suggest a more intimate connection between the concept of belief and the Christian tradition, and his endorsement of the substance of Needham's critique of the concept pushes him toward this position. Moreover he seems to imply that merely focusing on questions of what people believe is implicitly Christianizing, arguing that it is only in light of the Christian tradition that we would ‘attach any importance at all to having or not having beliefs’ (Ruel 1982: 22).
The impression that Ruel regards the very use of the concept as implicitly Christianizing is strengthened by his example of a ‘loaded’ use of the term. The unsuspecting Western anthropologist he singles out is Evans-Pritchard, citing the opening sentence of the first chapter of Witchcraft, Oracles, and Magic among the Azande, ‘Azande believe that some people are witches and can injure them in virtue of an inherent quality’ (Evans-Pritchard 1937: 21). But if Evans-Pritchard's use of ‘believe’ is loaded, it is not clear what would count as an ‘unloaded’ use of the term. For although Evans-Pritchard does not fully declare what ‘believe’ means, it requires only a cursory reading of his text to see that he uses the expression ‘believe that’ interchangeably with the expressions ‘think that,’ ‘say that,’ and ‘are of the opinion that,’ while elsewhere he uses free indirect discourse and expresses Zande beliefs/thoughts in assertions presented in a Zande voice (Streeter 2023: 179). Often he shifts between these ways of presenting Zande thought within a passage, as in the following discussion of Zande attitudes toward dogs (Evans-Pritchard 1937: 50):
Witchcraft is sometimes found in dogs and is associated with various other animals and birds. The dog has a malicious way of looking at people and is so greedy that it resembles human beings, and on account of these and other deficiencies Azande think that dogs are often witches. Their opinion is said to have been sustained by a few cases in which dogs have been proved guilty of witchcraft by judgements of the poison oracle.
Evans-Pritchard's text is an inevitable reference point for any anthropological discussion of belief, but it does not serve Ruel's purposes. For while it is easy to see why anthropologists and historians might worry about the Christianizing implications of the ascription of beliefs, it is hard to sustain this concern about the ascription of ‘thoughts’ or ‘sayings.’
What Ruel fails to appreciate is that he and Evans-Pritchard have different conceptions of belief ascription. For Ruel, the model is one of translation, or at least this seems implicit in his call for anthropologists ‘to be aware of the complexity of the concepts that we draw from our own culture, which have a history and contextual compulsion of their own which often ill-match the ideas and actions they are used to interpret’ (Ruel 1982: 9; my emphasis). Evans-Pritchard, however, did not understand himself to be matching a native idea when he ascribed beliefs to the Azande (Evans-Pritchard 1937: 70–71). He ascribed beliefs on the basis of what Azande said and did, and took them to believe p when they asserted p (or something that implied p) and reasoned and acted in a way that showed some commitment to the implications of p. Translation is involved in understanding what they say, but Evans-Pritchard was not translating anything in saying that the Zande believe ‘that some people are witches and can injure them in virtue of an inherent quality,’ just as he was not translating a native concept of thought in saying that Azande ‘think that dogs are often witches.’
The difference between Ruel and Evans-Pritchard is substantive, and has implications for any attempt to write the history of belief itself. For there can be no such history if Evans-Pritchard is right, since on his account the ascription of belief occurs at a very basic level of interpretation, that of understanding whether people mean what they say. Ruel's historicist concerns can only get off the ground if he can show that Evans-Pritchard misused the terms ‘believe’ and ‘belief.’ My own view is that this cannot be done, and that Ruel's translational conception of belief ascription has no grounding either in the ordinary language use of the terms ‘believe’ and ‘belief’ or in their use by anthropologists. This would tend to suggest that belief does not ‘have a history’ in the robust sense that some anthropologists and historians have claimed.
Conceptions of Belief Within History and Anthropology
If Ruel's failure to engage with Evans-Pritchard's conception and practice of belief ascription were based on a simple misunderstanding it would have no wider significance. But the conception of belief that guides Ruel's essay is widely shared among anthropologists who have reflected upon the concept. The difference between Ruel and Evans-Pritchard is therefore a token of a broader discrepancy between anthropologists’ reflective conception of belief and belief ascription and their historical practice of ascribing beliefs, Evans-Pritchard's use of the terms ‘believe’ and ‘belief’ being consistent with their use by roughly contemporary and slightly younger social anthropologists (Streeter 2023: 182-83).
Where there is such a discrepancy, it is the reflective conception that must be modified, and to do this we must understand how it came to be held. This conception has two key components, which have never been defended, and which tend to find expression in implicit and oblique ways. They are (i) that belief is a feeling/experience, and (ii) that the ascription of belief centrally involves the translation of a native concept of belief. I will address these assumptions in turn.
(a) Belief as an experience/feeling
The concerns that anthropologists have about the concept of belief were inaugurated by Needham's 1972 monograph Belief, Language, and Experience, and supplemented by the work of authors like Wilfred Cantwell Smith, Jean Pouillon, and Ruel. Needham's critique of the concept started from the premise that the word ‘belief’ ‘denotes’ an ‘inner state,’ and while he never satisfactorily articulated the meaning of this expression, he clearly had in mind a feeling like fear or anger (Needham 1972: 3). As we have seen, Ruel implicitly reiterated this conception of belief, and other anthropologists have done likewise (Pina-Cabral 2017: 82-86). More recently, and independently of anything in Needham, Ethan Shagan has endorsed a similar position, drawing the thought that belief might have a history from research into the history of emotions (Shagan 2018: 26–27).15
In addition to being explicitly classified as a feeling, there are occasions where belief is implicitly understood on this analogy. One of the more frequently recurring is the notion that belief is self-reflexive, and that in ascribing beliefs we are saying something about how people understand themselves. Thus Donald Lopez Jr. speaks of how the ‘accumulated weight’ of Christian discourse about belief has ‘resulted in the generally unquestioned assumption that adherents of a given religion, any religion, understand that adherence in terms of belief’ – this assumption apparently being implicit in the tendency of scholars of religion to focus on beliefs (Lopez 1998: 21, my emphasis). Likewise Joel Robbins worries that the ascription of beliefs to people from areas ‘outside the historical sphere of Christian dominance’ might be ethnocentric because ‘it is probably only moderns, and perhaps only elite moderns, who have understood themselves primarily to be engaged in believing that certain propositions are true about the world,’ while Jonathan Mair can argue that belief is often unimportant for anthropologists because ‘many religious people are not interested in beliefs’ (Robbins 2007: 14; Mair 2013: 451, my emphases).
That a move has been made in understanding belief ascriptions in this way I will try to show in the next section. But Lopez Jr., Robbins, and Mair are talking about something different from belief as it is understood by Evans-Pritchard. This becomes clear if we rephrase the opening sentence of the first chapter of Witchcraft, Oracles, and Magic as ‘Azande think that some people are witches and can injure them in virtue of an inherent quality.’ For this does not imply that Zande understand themselves to be engaged in thinking that some people are witches and can injure them in virtue of an inherent quality. Likewise the importance of this part of Zande thought would not be contingent on the importance the Azande self-consciously attach to it as something they think (we would not demand native utterances of the form ‘I think that some people are witches and can injure me in virtue of an inherent quality’ to have grounds either for ascribing this thought to them or for treating it as important).
What Lopez Jr., Robbins, and Mair have done, I think, is aligned belief with certain reflexive feelings which, because they ‘are themselves attitudes, modes of understanding the world as well as oneself,’ can only be experienced by people who possess the concept of the feeling. As Richard Moran has argued, ‘someone who cannot distinguish shame from embarrassment (or better, someone who recognizes no such distinction in principle, seeing only various forms of discomfort) will not be our best candidate for either emotion’ (Moran 2001: 41). Moreover, in such cases the anthropologist's estimation of the emotion's significance will be related to its importance to the people he studies (judged, for example, by the frequency with which they talk of their shame or embarrassment, or whatever the emotion happens to be).
As to how anthropologists have reached this position, there are several possibilities, but I think we can understand it best as proceeding from a particular conception of the meaning of the word ‘belief’ (cf. Wittgenstein 2009a: §90). Needham started from the assumption that ‘belief’ functions as a label for an ‘inner state’, and while subsequent contributors to anthropological debate about belief have not stated this premise so explicitly, there are hints that something like it guides their sense of the problems that surround the concept.
For if ‘belief’ were a label for a state of mind, several implications would follow. One is that first person and third person uses of the verb ‘believe’ would be symmetrical, with both referring to the subject's state of mind. And one way to develop this thought would be to take the first person present tense use of the verb as the key to the state, and to understand the ascription of belief to another to involve the ascription of a state of mind modelled on the information we often (although by no means invariably) convey with the first person use of the verb, as though whenever we say that people believe p we imply that their relation to p is that of someone who says ‘I believe p’. It is this assumption that underlies the notion of belief as a conviction that always includes doubt (Pouillon 1979: 48), or the conception of a belief as ‘a view to which we are reluctant to commit – a view we can't quite bring ourselves to take entirely seriously’ (Holbraad & Pedersen 2017: 192).
Of course ‘believe’ and ‘belief’ can be used to express doubt or doubtfulness. The question is whether the connection between belief and doubt is such that ‘belief’ strictly implies doubt, as Pouillon and Holbraad and Pedersen suppose. Here it is worth recalling that the verb ‘think’ can also be used to express doubt. If someone were to ask me whether Macron is president of France and I answer ‘I think so,’ I imply that he is but that I am not sure (‘I think’ here – as in many contexts – has the same use as ‘I believe’). But it does not seem plausible that ‘think’ entails doubt or doubtfulness, as though whenever we say that people think p we imply that they are reluctant to commit to p. Yet it is this inference that Pouillon and Holbraad and Pedersen invite us to draw about belief. I will suggest shortly that their argument involves a misunderstanding of our own ordinary practice of using the words ‘believe’ and ‘belief.’
(b) Ascription of belief as translation
While the picture of belief as a reflexive feeling encourages the thought that the ascription of belief is centrally a matter of translation, I suspect that this analogy is not the only thing that has led scholars to this understanding of belief ascription. For historians and anthropologists are habituated to worrying about the ethnocentric or anachronistic implications of their terms. Moreover, the centrality of translation within anthropological and historical interpretation has led some to think of cross-cultural interpretation as itself an exercise in translation (Lienhardt 1954: 97; Evans-Pritchard 1962: 22; cf. Asad 1993: 171–99; Ginzburg 2012: 109–10), and influential theorists of interpretation have given independent defences of this position (Gadamer 1989: 387).
The idea that we are translating something when we ascribe beliefs to people – ‘matching’ a native idea, to use Ruel's formulation – is therefore likely to be an intuitively compelling one for anthropologists and historians, and it is perhaps for this reason that no one has defended it. For the most part it is taken for granted, implicit in the concern that, in the absence of a native term that we can translate as ‘believe,’ we have no grounds for ascribing beliefs to people (Needham 1972: 1; Pouillon 1979: 46–48; Davies 2011: 413–14; Holbraad 2012: 63; van der Veer 2016: 44), or in the thought that it would be mistaken to ascribe beliefs in our sense to e.g. people in the middle ages, because the word ‘belief’ then meant something different from what it means now (W.C. Smith 1977: 41–44; 1979: 77, 143–50; 1981: 92–93; Robbins 2007: 14; Handelman 2008: 191; Asad 2012: 46–47; Harrison 2015: 49).
Conclusion: the Grammar of Belief
The conception of belief that I have sketched shapes debate about belief among anthropologists, but is not adequate to the way anthropologists have used the terms ‘believe’ and ‘belief.’ This gives us strong reasons for disavowing it, but we may supplement this conclusion by drawing attention to features of what Wittgenstein called the ‘grammar’ of the ordinary language use of the term ‘believe,’ particularly the distinctive character of the expression ‘I believe,’ which support Evans-Pritchard's account of the grounds upon which he ascribed beliefs to the Azande.
Wittgenstein's most sustained reflections on belief focus on what he called ‘Moore's paradox,’ or statements of the form ‘p but I don't believe p’.16 If someone were to tell us in all seriousness that ‘London is in England but I don't believe that London is in England,’ or ‘Macron is president of France but I don't believe Macron is president of France’ we would think that this person does not know how to use the word ‘believe.’ Yet the statements are not logically contradictory, since ‘p’ and ‘I believe p’ are different states of affairs, and it could be that Macron is president of France and I don't believe that Macron is president of France. Why, then, can I not say this myself?
Cora Diamond has noted how Wittgenstein's discussions of belief are concerned with ‘apparent anomalies in the concept, ways in which the concept apparently diverges from a schematism through which we might try to represent it’ (Diamond 2005: 99). One such schematism is Needham's conception of belief as a putative inner state, and Wittgenstein's discussions of Moore's paradox address an ancestor of this position.17 They show that the expression ‘I believe’ does not function in the way that this conception would lead us to expect, for Moore's paradox would not arise if ‘I believe p’ referred to my feeling and ‘p’ referred to the world (there is, for example, nothing absurd about the statement ‘it's raining but I don't fear that it's raining’). The absurdity of ‘p but I don't believe that p’ is a function of the fact that ‘p’ and ‘I believe p’ have roughly the same use, making the statement effectively, if not formally, a contradiction (Wittgenstein 2009b: §89). Wittgenstein acknowledges that the expression ‘I believe’ ‘throws light on my state’ insofar as it allows inferences to be drawn about my conduct, and he concedes that ‘there is a similarity here to manifestations of emotion, of mood, and so on.’ But he immediately adds that ‘if . . . “I believe it is so” throws light on my state, then so does the assertion “It is so.” For the sign “I believe” can't do it, can at the most hint at it’ (Wittgenstein 2009b: §96–97).
Having resolved the initial appearance of paradox of ‘p but I don't believe p,’ Wittgenstein explores a number of other seeming irregularities that Moore's phrase brings to light, particularly the asymmetry between first person and third person uses of the verb ‘believe.’ For Moore's paradox only arises in the first person present tense: ‘Macron is president of France but he doesn't believe that Macron is president of France’ is perfectly intelligible, as is ‘Macron was president of France but I didn't believe that Macron was president of France.’ Moreover, the near synonymy of ‘p’ and ‘I believe p’ does not hold when we embed ‘I believe p’ in a hypothetical clause: ‘suppose I believe Macron is president of France’ is not roughly synonymous with ‘suppose Macron is president of France’ (Wittgenstein 2009b: §87; 1980a: §478). Wittgenstein explains these seeming irregularities by situating the use of ‘believe’ within the different attitudes that we take towards ourselves and toward others (Wittgenstein 2009b: §103). I ascribe beliefs to others on the basis of what they say and do. The question can arise of how I know that they believe p, and there is the possibility of my mistakenly ascribing the belief to them (it may be that they merely seem to believe p). But the expression ‘I believe p’ is not based on observation of myself and is therefore also not something I ascribe to myself, as though it were a description or a report (Wittgenstein 2009b: §99; cf. 1980a: §504). Rather it is, standardly, a particular expression of the belief that p, an avowal of the belief that I express more directly with the simple assertion ‘p’.
For Wittgenstein, then, the expression ‘I believe’ is what William Child has nicely summarized as a ‘learned addition to a more basic use of language which expresses our beliefs’ (Child 2011: 173–74). A corollary is that most of the language that expresses our beliefs does not involve the use of the word ‘believe,’ for we express our beliefs most directly in simple assertions. We use the phrase ‘I believe p’ when we reflect on our beliefs, and we can use it to express our beliefs in a qualified way, rather as we might use ‘I intend’ to qualify an expression of intention or ‘I would like’ to qualify the expression of a wish (compare ‘I intend to go to Paris tomorrow’ with ‘I'm going to Paris tomorrow,’ or ‘I'd like a glass of wine’ with ‘wine over here’) (Williams 1973: 138; Wittgenstein 1980a: 472). But these uses of the verb presuppose the mastery of the practice of assertion, which is the practice of expressing our beliefs.
If Wittgenstein's account of the grammar of belief is right, then what I have called the conception of belief within anthropology is mistaken: belief is not a feeling or experience, and the ascription of belief should not be understood on the model of translation. Absent these assumptions, it is hard to see how we can form a conception of belief as historically and culturally contingent in the sense that its ascription could be anachronistic or ethnocentric. For if the most direct expression of the belief that p is the assertion p, as Evans-Pritchard assumed and Wittgenstein's discussions of the ‘grammar’ of belief showed, then the ascription of belief is involved at a fundamental level of interpretation, for we understand people as expressing what they believe whenever we take them to mean what they say. If belief can be said to have a history, then, it can only have one in a modest sense.
Acknowledgements
My thanks to John Arnold, Dana Fields, John North, Robert Parker, and Phil Swift for helpful comments and encouragement on earlier drafts of this article.
Notes
Street (2010: 260); cf. Pouillon (1979: 49–51); Price (1984: 10–11); Asad (1993: 47–48); Lopez, jr. (1998: 33); Davies (2011: 397); Holbraad (2012: 63); Mitchell (2017: 331). Fenella Cannell notes that many anthropologists have avoided the ‘category of belief’ owing to its ‘Christocentric bias,’ although she does not endorse or criticize this tendency; Cannell (2010: 97); cf. Coleman (2010: 799).
The locus classicus for this view is W.C. Smith (1977: 36–69; 1979: 105–17; 1981: 92–93). Anthropological discussions indebted to his work include Good (1994: 15–17); Robbins (2007: 14); Glazier (2008: 27–28); Handelman (2008: 191); Vilaça (2016: 161).
An important text that equivocates between these claims is Robbins (2007: 14).
The newspaper examples are drawn from The Guardian.
‘Category’ can refer to a concept's extension, in which case a history of the category of belief would just be a history of belief. But as I read him, Shagan treats ‘category,’ ‘concept,’ and ‘conception’ as synonyms.
Shagan works back from a set of intuitions or second-order reflections about belief, and wants in particular to historicize ‘the sense that belief is synonymous with private judgement, and therefore modern subjects believe or disbelieve according to their own conception of whether a given proposition is credible’: Shagan (2018: 6). But this is more like a theory of belief than the concept of belief.
Shagan never satisfactorily contextualizes debates about Christian belief in relation to more mundane uses of the term ‘believe,’ even though he concedes that it ‘referred’ in every European language ‘both to religion and to more banal truth-claims, because no alternative could capture the ineffable qualities that made religious belief different from ordinary intellective assent’ (Shagan 2018: 43). But if this is so, there is a question about what the various attempts to redefine the nature of Christian belief could have achieved. Shagan tends to imply that the reformers’ claims were self-establishing, as though it were enough to assert that belief is difficult for belief to become difficult.
Harris (2001: 25). There is a large bibliography on ancient anger, but see further the essays in Braund and Most (2002) and Burnyeat (2022: 356–376).
Goodstein implies that this question should be avoided, and that rather than trying to define ‘boredom’ (which she thinks impossible) we should attempt ‘a genealogy of the experience’ (Goodstein 2005: 12; cf. 2020: 42–43). It may be right that boredom cannot be defined in terms of necessary and sufficient conditions, but it does not follow that the question of the definition of boredom – the question of what boredom is – is misconceived, and there are several ways in which it may be addressed besides a formal, theoretical, definition. Indeed, Goodstein's study itself contains several descriptions of the phenomenology of boredom, which guide her attempt to historicize the experience (cf. Goodstein 2005: 1–2, 18). I do not see how scholars interested in the history of boredom can avoid addressing the question of what boredom is, and I cannot see how they can answer it by way of a genealogy, for they must have some understanding of what the experience is if they are to trace its genealogy.
Ron Mallon has noted that ‘opulent construals of the meaning of terms can guarantee that provocative claims of cultural locality will turn out to be true’ (Mallon 2016: 53) and I suspect this applies to Goodstein's argument. At least, her characterization of boredom seems to me highly scholastic – consider her claim that there is a ‘nihilistic dynamic internal to the experience’ (Goodstein 2005: 23), that ‘even in its most quotidian manifestations . . . boredom embodies a specifically modern crisis of meaning’ constituted by the loss of earlier ‘certainties of faith, tradition, sensation,’ the modern ‘form of human existence keyed to the precision of what Georg Simmel called the “supersubjective temporal schema” of clock time,’ and the new ‘ideal of human life itself as a process of incessant change and improvement’ (Goodstein 2005: 3-6). But can it be true that children complaining of boredom on a long car journey are giving voice to an experience with that content and philosophical significance? Or should we say that these children do not know the meaning of the word ‘boredom’ and are not really bored? Goodstein would have to make good one or other of these claims, neither of which looks plausible. My suspicion is that her argument for boredom's modernity depends on a series of conflations – of a literary cum philosophical ‘discourse on boredom’ with the experience of boredom, and of the meaning of the word boredom with its various literary and philosophical connotations. All of which is not to say that boredom is not a distinctively modern experience, only that Goodstein does not do enough to establish this claim.
Paul Veyne's notion of ‘modalities of belief’ is a descendant of James’ position (mediated by Alfred Schutz); Veyne (1988: xi, 27–39, 135–36 n33). I find Veyne's discussion of belief unclear, and I have for this reason not discussed it in the body of the text. There is a helpful attempt at articulating his notion of ‘modalities of belief’ at Mair (2013: 458–59), but Mair ascribes a more consistent position to Veyne than the text supports. For Veyne never clearly answers his own question, but sometimes asserts that the Greeks believed in their myths, while elsewhere he implies that their attitude was something other than belief, as when he suggests that, for contemporaries of Pindar or Homer, ‘Myth was a tertium quid, neither true nor false’ (28).
Wilfred Cantwell Smith contested this claim (for references, see n2), but he ignored the most important evidence for the ‘religious importance’ of beliefs to pre-modern Christians, namely the recurring conflicts among Christians over questions of orthodoxy. This is an extraordinary omission, and the implausibility of his historical narratives (which imply that questions of what people should believe were ‘religiously unimportant’ for Christians during the Reformation) should have led him to think about where he might have gone wrong.
On the genesis of the Christian concept of heresy, see Le Boulluec (1985: 110–12).
As this will suggest, I think questions about the explanatory importance of belief should be addressed on a case-by-case basis, but I accept that the accounts people offer of their own practices or institutions are not always probative and that anthropologists or historians may justifiably go beyond or even ignore native beliefs in explaining these practices and institutions. There may also be certain kinds of social phenomena that cannot be fully understood in terms of their participants’ beliefs (as, for instance, has been argued of ritual by Stanley Tambiah and Roy Rappaport). But these arguments are different from arguments against the use of the concept of belief on the grounds of ethnocentrism or anachronism.
Shagan is not alone in taking inspiration from this source, but I do not see how one can make sense of emotions as historically and culturally contingent without the concept of belief (implying that belief itself is not historically contingent in the same manner or degree as the emotions); see Shweder (2004: 83 n.1), remarking that ‘Beliefs . . . should . . . be part of any theoretical language for comparative research on mental states.’
The central text is section 10 of Wittgenstein (2009b), but there is supplementary material in Wittgenstein (1980a) and (1980b). This is the most important part of Wittgenstein's work on belief, and the one for which the textual evidence is most reliable. The evidence for his other views on the subject is of mixed quality, and the texts collected as his discussions of religious belief are especially problematic; see Mulhall (2001: 96).
Needham claimed to be addressing Wittgenstein's question: ‘Is belief an experience?’, which alludes to Hume's conception of belief as a ‘feeling or sentiment’ (Needham 1972: xiii; Hume (1738) 2007: 68; Munz & Ritter 2017: 204-10). Wittgenstein's discussions of belief engage more directly with William James’ Humean conception of belief (James (1890) 1981: 913–16).
Bibliography
Asad, Talal. 1993. Genealogies of Religion: Discipline and Reasons of Power in Christianity and Islam. Baltimore: Johns Hopkins University Press.
Asad, Talal. 2012. “Thinking about religion, belief, and politics.” In The Cambridge Companion to Religious Studies, ed. Robert Orsi, 36–57. Cambridge: Cambridge University Press.
Brown, Peter. 2012. Through the Eye of a Needle: Wealth, the Fall of Rome, and the Making of Christianity in the West, 350–550 AD. Princeton: Princeton University Press.
Burnyeat, Myles. 2022. Explorations in Ancient and Modern Philosophy, Volume 3. Cambridge: Cambridge University Press.
Cannell, Fenella. 2010. “The anthropology of secularism.” Annual Review of Anthropology 39: 85–100.
Child, William. 2011. Wittgenstein. Abingdon: Routledge.
Coleman, Simon. 2010. “An anthropological apologetics.” South Atlantic Quarterly 109.4: 791–810.
Davies, Jason. 2011. “Believing the evidence.” In Evidence, Inference and Enquiry = Proceedings of the British Academy 171, ed. P. Dawid, W. Twining, and M. Vasilaki, 395–434. Oxford: Oxford University Press.
Diamond, Cora. 2005. “Wittgenstein on religious belief: the gulfs between us.” In Religion and Wittgenstein's Legacy, ed. D.Z. Phillips and M. von der Ruhr, 99–137. London: Routledge.
Evans-Pritchard, Edward E. 1937. Witchcraft, Oracles, and Magic among the Azande. Oxford: Clarendon Press.
Evans-Pritchard, Edward E. 1962. Essays in Social Anthropology. London: Faber & Faber.
Febvre, Lucien. (1942) 1982. The Problem of Unbelief in the Sixteenth Century: The Religion of Rabelais, translated by B. Gottlieb. Cambridge MA: Harvard University Press.
Gadamer, Hans-Georg. 1989. Truth and Method, 2nd edition. Trans. Joel Weinsheimer and Donald G. Marshall. London: Sheed & Ward.
Ginzburg, Carlo. 2012. “Our words and theirs: a reflection on the historian's craft, today.” In Historical Knowledge: In Quest of Theory, Method and Evidence, ed. S. Fellman and M. Rahikainen, 97–119. Cambridge: Cambridge Scholars Publishing.
Glazier, Stephen. 2008. “Demanding deities and reluctant devotees: belief and unbelief in the Trinidadian Orisa movement.” Social Analysis 52.1: 19–38.
Good, Byron J. 1994. Medicine, Rationality, and Experience: An Anthropological Perspective. Cambridge: Cambridge University Press.
Goodstein, Elizabeth. 2005. Experience Without Qualities: Boredom and Modernity. Stanford: Stanford University Press.
Goodstein, Elizabeth. 2020. “Boredom and the disciplinary imaginary.” In The Culture of Boredom, ed. J.R. Velasco, 23–54. Leiden: Brill.
Hacking, Ian. 1999. The Social Construction of What? Cambridge MA: Harvard University Press.
Handelman, Don. 2008. “Afterword: returning to cosmology. Thoughts on the positioning of belief.” Social Analysis 52.1: 181–95.
Harris, William V. 2001. Restraining Rage: The Ideology of Anger Control in Classical Antiquity. Cambridge MA: Harvard University Press.
Harrison, Peter. 2015. Territories of Science and Religion. Chicago: University of Chicago Press.
Holbraad, Martin. 2012. Truth in Motion: The Recursive Anthropology of Cuban Divination. Chicago: University of Chicago Press.
Holbraad, Martin, and Morten Axel Pedersen. 2017. The Ontological Turn: An Anthropological Exposition. Cambridge: Cambridge University Press.
Hume, David. (1738) 2007. A Treatise of Human Nature: A Critical Edition. Volume 1: Texts, ed. D.F. Norton and M.J. Norton. Oxford: Oxford University Press.
James, William. (1890) 1981. The Principles of Psychology. Cambridge MA: Harvard University Press.
Leary, David E. 2018. The Routledge Guidebook to James's Principles of Psychology. London: Routledge.
Le Boulluec, Alain. 1985. La notion d'hérésie dans la littérature grecque, IIe-IIIe siècles I: De Justin à Irénée. Paris: Études augustiniennes.
Lienhardt, Godfrey. 1954. “Modes of thought.” In The Institutions of Primitive Society: A Series of Broadcast Talks, ed. Edward E. Evans-Pritchard, 95–107. Oxford: Oxford University Press.
Lopez, Jr., D.S. 1998. “Belief.” In Critical Terms for Religious Studies, ed. M.C. Taylor, 21–35. Chicago: University of Chicago Press.
Mair, Jonathan. 2013. “Cultures of belief.” Anthropological Theory 12.4: 448–66.
Mallon, Ron. 2016. The Construction of Human Kinds. Oxford: Oxford University Press.
Mitchell, Jon P. 2017. “Defining religion: Geertz and Asad.” In Religion, Theory, Critique: Classic and Contemporary Approaches and Methodologies, ed. Richard King, 327–34. New York: Columbia University Press.
Moran, Richard. 2001. Authority and Estrangement: An Essay on Self-Knowledge. Princeton: Princeton University Press.
Mulhall, Stephen. 2001. “Wittgenstein and the philosophy of religion.” In Philosophy of Religion in the 21st Century, ed. D.Z. Phillips and T. Tessin, 95–118. New York: Palgrave.
Munz, Volker A. and Bernhard Ritter, eds. 2017. Wittgenstein's Whewell's Court Lectures Cambridge, 1938–1941, From the Notes by Yorick Smythies. Oxford: Wiley Blackwell.
Needham, Rodney. 1972. Belief, Language, and Experience. Oxford: Blackwell.
Pina-Cabral, João de. 2017. World: An Anthropological Examination. Chicago: Hau Books.
Pouillon, Jean. 1979. “Remarques sur le verbe “croire”.” In La Fonction symbolique: essais d'anthropologie, ed. M. Izard and P. Smith, 43–51. Paris: Gallimard.
Price, Simon R.F. 1984. Rituals and Power: The Roman Imperial Cult in Asia Minor. Cambridge: Cambridge University Press.
Robbins, Joel. 2007. “Continuity thinking and the problem of Christian culture: belief, time, and the anthropology of Christianity.” Current Anthropology 48.1: 5–17.
Ruel, Malcolm. 1982. “Christians as believers.” In Religious Organization and Religious Experience, ed. J. Davis, 9–31. London: Academic Press.
Shagan, Ethan. 2018. The Birth of Modern Belief: Faith and Judgement from the Middle Ages to the Enlightenment. Princeton: Princeton University Press.
Shapin, Steven. 1994. A Social History of Truth: Civility and Science in Seventeenth-Century England. Chicago: University of Chicago Press.
Sheehan, Jonathan. 2010. “When was disenchantment? History and the secular age.” In Varieties of Secularism in a Secular Age, ed. M. Warner, J. Van Antwerpen and C. Calhoun, 217–42. Cambridge MA: Harvard University Press.
Shweder, Richard. A. 2004. “Deconstructing the emotions for the sake of comparative research.” In Feelings and Emotions: The Amsterdam Symposium, edited by A. S. R. Manstead, N. Frijda, and A. Fischer, 81–97. Cambridge: Cambridge University Press.
Smith, Wilfred Cantwell. 1977. Belief and History. Charlottesville: University of Virginia Press.
Smith, Wilfred Cantwell. 1979. Faith and Belief. Princeton: Princeton University Press.
Smith, Wilfred Cantwell. 1981. Towards a World Theology: Faith and the Comparative History of Religion. London: Macmillan.
Smith, William Robertson. 1894. Lectures on the Religion of the Semites I: The Fundamental Institutions, 2nd edition. London: Adam and Charles Black.
Street, Alice. 2010. “Belief as relational action: Christianity and cultural change in Papua New Guinea.” Journal of the Royal Anthropological Institute 16.2: 260–78.
Streeter, Joseph. 2023. “‘Belief’ and anthropology, in use and in theory.” Method and Theory in the Study of Religion 35: 174–200.
Taylor, Charles. 2007. A Secular Age. Cambridge, MA: Harvard University Press.
van der Veer, Peter. 2016. The Value of Comparison. Durham NC: Duke University Press.
Veyne, Paul. 1988. Did the Greeks Believe in their Myths? An Essay on the Constitutive Imagination, translated by P. Wissing. Chicago: University of Chicago Press.
Vilaça, Aparecida. 2016. Praying and Preying: Christianity in Indigenous Amazonia. Berkeley: University of California Press.
Williams, Bernard. 1973. Problems of the Self. Cambridge: Cambridge University Press.
Wittgenstein, Ludwig. 1980a. Remarks on the Philosophy of Psychology I. Oxford: Basil Blackwell.
Wittgenstein, Ludwig. 1980b. Remarks on the Philosophy of Psychology II. Oxford: Basil Blackwell.
Wittgenstein, Ludwig. 2009a. Philosophical Investigations, Revised 4th edition, edited by Peter M.S. Hacker and Joachim Schulte. Oxford: Wiley-Blackwell.
Wittgenstein, Ludwig. 2009b. “Philosophy of psychology – A fragment.” In Wittgenstein, Philosophical Investigations, Revised 4th edition, ed. Peter M.S. Hacker and Joachim Schulte, 183–243. Oxford: Wiley-Blackwell.