By Barry Hallen

Barry Hallen, Morehouse College and W.E.B. Du Bois Institute, Harvard University

More than three decades ago a definite and deliberate consensus had been reached among philosophers in and of Africa that there had been enough second-order talk about whether there had been and was ‘philosophy’ in indigenous African cultural contexts. This meant that it was time to move on, and to produce individualized, specialized philosophical studies that arose from those cultural contexts. There was also a consensus that to ‘translate’ internationally in scholarly terms this material needed to be presented or at least introduced in and by formats that would not be entirely foreign to non-African audiences.

In my own case, I settled on an adaptation of what has come to be known in the Western tradition as ordinary language philosophy. One good reason for this choice was that it made sense, in practical terms for someone like myself who is not an African and who was coming to terms with a language culture that was not my own, to adopt an approach that essentially involved learning more about the language as well as the culture. Also, the ordinary language approach is implicitly conceptually oriented insofar as concentrating on correct and incorrect usage can involve identifying the criteria governing the usage of words that are targeted because of potential philosophical prepossession.

Another central interest I had at the time was to put W.V.O. Quine’s indeterminacy thesis of radical translation to some form of empirical test—particularly with regard to that aspect of the thesis he discusses under the heading of “radical translation.”1

I found the open-mindedness of the thesis regarding the possibility of radically different meanings in radically different languages liberating. To someone who had been conditioned to regard paradigms espoused by English- language analytic philosophy as normative, it opened the door to the possibility of fundamentally different meanings of the terminology used in everyday, conventional situations—with their own intrinsic rational integrity—existing in languages that might not have reason to share a single cognate in common with English. This of course underlined the possibility that there might be theoretical networks intrinsic to Africa’s language cultures that, on philosophical grounds, had not received their just deserts.

Last, but far from least, the ordinary language approach gave me, as someone whose mother discipline was philosophy rather than anthropology, a methodology of philosophical origin that would entitle me to circumvent the rather loud objections of social anthropological colleagues who insisted that I had no business undertaking any form of research that involved anything resembling fieldwork.

Those introductory points noted, let me proceed to the topic of ethical knowledge. The title does not imply, as perhaps some might think, that this essay’s central focus will involve how it may be possible to establish ethical principles or moral values on some sort of secure objective basis, however tantalizing that prospect may remain. Indeed the dominance of that presumption, if anything, is one indication of how important it is to suspend the importation of pre-existing concerns when coming to terms with a ‘radically’ different language culture’s conceptual system(s).

Abstract thought in Africa’s indigenous cultures has been said to be expressed primarily in and by myths, stories, proverbs, and rituals. Comparative studies of abstract thought in the West and Africa supposedly suggest there is little point in asking whether its African forms are ‘true’ because, even if believed so by members of the relevant cultures, they are too obviously ‘fictions’ and ‘exercises’ invented to permit Africans at least to feel that they understand and thereby can exercise some control over the forces underlying life’s sometimes paradoxical events. As such they are said to fulfill people’s emotional needs as much or more than preeminently intellectual ones.

Many African intellectuals have protested that critical reasoning also has to play an essential role in African systems of thought and that, in any case, dividing the person up between a rational self and an emotional self is an hypothesis of Western cultural orientation.2 Other African scholars (Wiredu 1980) have suggested that lumping all of African ‘abstract’ thought into a single category and then comparing it with the theories of so deliberately and painstakingly refined a subject as academic philosophy is not fair.3 Africa has its own folklore, folk thought or folk philosophy (relatively popular beliefs, superstitions, etc.), as is also the case with the West. This distinction in the two cultures’ respective modes-of-thought or beliefs has first to be made clear before a neutral basis for intercultural comparisons will be established. Otherwise, elements of African folklore might end up being contrasted and compared (unfavorably, of course) with technically abstract, disciplinary-specific methodologies and theories. Indeed the disinterest in things African on the part of Western analytic philosophers was almost certainly a consequence of the fact that they passively (or perhaps impassively), tacitly, had acquiesced to the portrait of Africa as a place where people did not assign a high priority to reason, to critical thinking, in formulating their views of the world. For one oddity about the analytic tradition as practiced within the academy is that virtually the only language that it has been used to analyze is English. The most obvious explanation for this is that analytic philosophy is a product of English-language culture. But was this really sufficient to explain why its method and techniques had never been applied in even an experimental manner to any non– Western language? Was there not here also evidence, albeit implicit, of a tacit judgment on the part of the Western academy that such endeavors were likely not to be philosophically rewarding?

In any case, the point of the Hallen-Sodipo research project on African Thought/Philosophy was to apply the techniques of analytic philosophy, as adapted for use in a culture that was substantially oral in nature,4 to the language of the Yoruba of southwestern Nigeria. Some of the Yoruba fields of discourse preselected as of particular interest were those relevant to epistemology or the theory of knowledge and ethics. It is the interrelations between elements of the Yoruba conceptual system relating to these two special interests that has resulted in the coining of the expression “moral epistemology.”

In Western epistemological theory the most problematic and controversial sub-category of information is what has come to be known as propositional knowledge. Generally this is associated with information in written or oral propositional (sentential) form that is supposed to be knowledge and therefore true, but which the individual recipient is in no position to test or to verify. When one reflects upon what a member of Western society may ‘learn’ in the course of a lifetime, it becomes clear that most people’s ‘knowledge’ consists of information they will never ever be in a position to confirm in a firsthand or direct manner. What they ‘find out’ from a history book, ‘see’ via the evening news on television or ‘confirm’ about a natural law on the basis of one elementary experiment in a high school physics laboratory—all could be (and sometimes are!) subject to error, distortion or outright fabrication.

Propositional knowledge is therefore generally characterized as secondhand, as information that cannot be tested or proved in a decisive manner by most people who have it and therefore has to be accepted as true because it ‘agrees’ with common-sense or because it ‘corresponds’ to or ‘coheres’ with the very limited amount of information that people are able to test and confirm in a firsthand or direct manner. Exactly how this coherence or correspondence is to be defined and ascertained is still a subject of endless wrangling in (Western) epistemological theory. What is relevant to the present discussion is that this wrangling is evidence of the intellectual concern and discomfort (in academic parlance it becomes one of the ‘problems’ of philosophy) on the part of (Western) philosophers about the weak evidential basis of so much of the information that people in that culture are conditioned to regard as knowledge, as true.

The distinction made in Yoruba-language culture between putative ‘knowledge’ and putative ‘belief’ reflects a similar concern about the evidential status of firsthand versus secondhand information. Persons are said to ‘know’ or to have ‘knowledge’ only of experience they have witnessed in a firsthand or personal manner. The example most frequently cited by discussants, virtually as a paradigm, is visual perception of a scene or an event as it is taking place. ‘Knowledge’ is said to apply to sensory perception generally, even if what may be experienced directly by touch is more limited than is the case with perception. ‘Knowledge’ in a Yoruba context implies a good deal more than mere sensation, of course. Perception implies cognition as well, meaning that persons concerned must comprehend that and what they are experiencing. The terms “ooto”/“otito” are associated with “knowledge” in certain respects that parallel the manner in which “true” and “truth” are paired with “know”/“knowledge” in the English language. In the English language “truth” is principally a property of propositional knowledge, of statements human beings make about something, while in Yoruba “ooto” may be a property of both propositions and certain forms of experience. Therefore in some contexts it is better rendered into English as being “certain” or “certainty.”

The Yoruba noun form that I am rendering as ‘belief’—“igbagbo” (and its verb form “gbagbo”)—does in fact arise from the conflation of “gba” and “gbo.” The two components are themselves verbs, the former conventionally translated into English as “received” or “agreed to,” and the latter as “heard” or “understood.” Yoruba linguistic conventions suggest that treating this complex term as a synthesis of the English language “understood” (in the sense of cognitive comprehension) and of “agreed to” (in the sense of affirming or accepting new information one comprehends as part of one’s own store of secondhand information) is perhaps the best way to render its core meaning. Igbagbo encompasses what one is not able ‘to see for oneself’ or to experience in a direct, firsthand manner. For the most part this involves things we are told about or informed of—this is the most conventional sense of ‘information’—by others.

What makes it different from the English language “believe”/“belief” is that igbagbo can apply to everything that might be construed as secondhand information. This would apply to most of what in English-language culture is regarded as propositional knowledge: the things one is taught in the course of a formal education, what one learns from books, from other people and, of particular interest in the special case of the Yoruba, from oral traditions. While English-language culture decrees that propositional or secondhand information, since classified as ‘knowledge,’5 should be accepted as true, Yoruba usage is equally insistent that, since classified as igbagbo (putative ‘belief’), it can only be accepted as possibly true (o se e se) or untrue (ko se e se).

The cross-cultural ramifications of these differing viewpoints on the truth status of propositional or secondhand knowledge are worth considering. Yoruba-language speakers would likely regard members of English-language culture, who are willing to assign so much certainty to and put so much trust in information that they can never test or verify, as dangerously naive and perhaps even ignorant. While members of English-language culture might criticize their Yoruba counterparts’ identification of optimal knowledge with ‘you can only know what you can see’ as indicative of a people who have yet to discover the benefits of institutionalized knowledge and formal education.

The criteria that define the respective extents of and the interrelations between ‘knowledge’ (imo) and ‘belief’ (igbagbo) in Yoruba stipulate that any experience or information that is not firsthand, personal and direct must by definition fall under the heading of ‘belief’ (igbagbo). The sense of ‘belief’ (igbagbo) may therefore be paraphrased as “comprehending, and deciding to accept as possible (as ‘possibly true’ rather than as ‘true’), information that one receives in a secondhand manner.”

‘Knowledge’ (imo; firsthand experience) and ‘belief’ (igbagbo; information gained on the basis of secondhand experience) together exhaust all of the information that human beings have at their disposal. If and when my ‘knowledge’ (imo) is challenged by other persons who have not undergone a similar firsthand experience and who therefore doubt what I say I actually saw happen, the best way to convince them would be to arrange for some kind of test whereby they will be able to see the thing happen for themselves.6 If I cannot arrange for this kind of direct testing, the next best I can do is to ask any others who may have personally witnessed my own or a similar experience to come forward and testify. In this case my firsthand experience cannot become the challengers’ own ‘knowledge’ (imo), but if they are influenced by the combined testimony they may decide to ‘believe’ me/us and accept the information on a secondhand basis, as ‘belief’ (igbagbo).

A simple example may serve to clarify things. If I claim I have seen for myself (imo) that a certain friend drives a specific make and model of car and another friend challenges my claim, the best way to resolve the dispute is to visit the friend and see (imo) what kind of car she actually has. If the friend lives a thousand miles away, a more practical solution would be to ask other mutual friends who have seen (imo) the car themselves to tell us (igbagbo) what kind it is. Or perhaps to telephone my friend directly and ask her to tell us (igbagbo) what kind of car she is driving. Speaking to her directly by telephone still would not be firsthand ‘knowledge’ (imo) about the car because one is not actually seeing it. One is only hearing a further form of secondhand information about the car, another form of testimony—albeit a particularly relevant one given the circumstances.

If and when my ‘belief’ (igbagbo) is challenged by another person, again the best solution would be to arrange some form of empirical test. In this case since this is information I myself only know secondhand, the most reliable solution for all concerned would be to test it directly, so that the information would progress from being ‘belief’ (igbagbo) to being ‘knowledge’ (imo) for all concerned, myself included. Next best would again be to call upon all relevant witnesses who may have heard the same or similar secondhand information (igbagbo) or, even more definitively, have firsthand (imo) experience of what I can only claim to know on a secondhand (igbagbo) basis.

When agreement or a consensus among disputants is reached on the level of ‘belief’ (igbagbo), the applicable term (comparable to the role of “truth” with reference to knowledge, or of “ooto” with regards to imo) is “papo,” which may be rendered colloquially as ‘the words have come together.’7 The antecedent process of testimony, discussion and reflection on the basis of which the consensus is reached is described as “nwadi”—an expression whose meaning may be compared to the English-language ‘let’s get to the bottom of this matter.’8

The system that emerges from these criteria appears to be three-tiered. ‘Knowledge’ (imo) is the sole category of experience or of propositions entitled to be regarded as certain or as true (ooto). ‘Belief’ (igbagbo) that is in principle open to empirical testing, verification, and thereby transformed into ‘knowledge’ (imo that is ooto) is the next best. ‘Belief’ (igbagbo) that can never be verified and can only be evaluated on the basis of testimony, explanation, discussion and reflection (nwadi) is the least certain.

The significance of all this for cross-cultural understanding and comparisons is complex. The most obvious and perhaps important point is that Yoruba discourse does employ terminology and systematic criteria for the evaluation of any type of information. This is a priority to which African systems of thought were once said not to attach special importance or about which they were said to be unclear.

The moral underpinnings to this discussion of Yoruba epistemology become evident once one recognizes that the primary source of propositional or secondhand information in a culture that is significantly oral is other persons. For, if that is the case, knowledge of those other persons’ moral characters (iwa)—their honesty, their reliability as sources of information—becomes a fundamental criterion to evaluating the reliability of secondhand information obtained from them.

Knowledge of another person’s moral character is said to be obtained, most reliably, from observing (firsthand) their behavior (isesi). And in Yoruba discourse ‘behavior’ conventionally extends to ‘what they say’ and ‘what they do,’ which also pretty much corresponds to the standard Western notions of verbal and non-verbal behavior. But what is again in evidence here is the priority the Yoruba place upon hard evidence, upon only being able to ‘know’ what you witness in a firsthand manner. For the point is that a person’s verbal and non-verbal behavior are construed as firsthand evidence (imo) of their moral character (iwa).

Needless to say, a person’s moral character (iwa) is not as readily observable as everyday material objects, such as a tree or a table. Obviously a process of inference is involved in order to move from observing a multiplicity of individual actions to a generalization about character. But with specific reference to epistemological concerns—the person as a source of reliable secondhand information—the interplay between ‘knowledge’ (imo) and ‘belief’ (igbagbo) appears to be as follows. On the basis of a number of specific previous occasions when you have had the opportunity, firsthand (imo), to verify the truth (ooto) of a person’s statements, you are justified in using these firsthand experiences as the basis for a generalization about their moral character. This generalization may then serve as a kind of character reference for evaluating the reliability of future statements made by this same person but, strictly speaking, such evaluations must remain hypothetical or tentative until also confirmed in a firsthand manner.

What the overall process appears to involve is a kind of sliding scale for gauging varying degrees of epistemic certainty about the moral characters of and/or information provided by other persons. Those you have associated with directly and therefore have had ample opportunity to observe in a firsthand manner are those whose character you are in a position to know best, and thereby to judge whether information of which they are the source is likely to be reliable or unreliable. Those you have not associated with at all and therefore have had no opportunity to observe in a firsthand manner (or even to have heard at least something about in a secondhand manner) are those whose character you do not know, and thereby have no substantive basis on which to judge whether information of which they are the source is likely to be reliable or unreliable.

A person who makes an informative statement may be obliged to recount the precise circumstances in which he or she came by it. A person is expected to say whether there is any cause for uncertainty or imprecision about the information. Determining whether the information is derived from the speaker’s firsthand (imo) or secondhand (igbagbo) experience is part of this process. A person’s diligence in doing all of this also is considered important evidence of their moral character (iwa). With specific reference to what is here being characterized as a “moral epistemology,” at least four positive behavioral values are emphasized: (1) being scrupulous about the epistemological basis for whatever one claims to know, to believe, or to have no information about; (2) being a good listener, with the emphasis upon cognitive understanding rather than a polite and respectful demeanor; (3) being a good speaker, with the emphasis upon speaking in a positive, thoughtful, and perceptive manner rather than mere elocution; (4) having patience, with the emphasis upon being calm and self-controlled in judgment and intellect rather than merely in manner and demeanor.

The public in Western societies have become concerned about exercising control over the quality of information put out by the media. In an oral culture the media are people’s mouths. These four values, in effect, set broadcasting standards for those mouths. ‘Speaking well’ and ‘hearing well,’ as values, further reinforce the importance of providing accurate information or reliable advice and being forthright about the epistemological origins of that information and advice. A consciousness that cultivates ‘patience,’ especially in difficult or problematic situations, is more likely to maintain self-control and thereby optimal communication with its environment. ‘Speaking well’, ‘hearing well,’ and ‘patience’ are not, then, moral values in any conventional sense. They are rather epistemological virtues because of their instrumental value for promoting the accuracy of information.

One interesting philosophical consideration about this Yoruba alternative epistemology would be its consequences for the conventional definition of propositional knowledge as arising from ‘justified, true belief.’ Obviously the criterion of belief would no longer apply, since it would be pointless and confusing to say that one had secondhand, possibly true information about what one already ‘knew’ to be true on the basis of firsthand experience. Also, the notion of justification would seem to lose much of its significance since one is not obligated to provide the kinds of evidence associated with that criterion for information or experience that already has been validated in a firsthand manner.

All of this also could have substantive ramifications for the so-called Gettier counter- examples, since all or many of them would be de-fanged by the Yoruba stipulation that what is said to be ‘known’ must have been witnessed in a firsthand manner throughout. And last, but far from least, these distinctive and different criteria for distinguishing expressions such as ‘I know that’ from ‘I believe that’ could justify the claim that the attributes of so-called propositional attitudes (“I know that,” “I hope that,” “I fear that,” etc.) may be relative to particular language cultures (hence the use of ‘fools’ quotation marks’ throughout my text).


Works Cited

Abraham, R.C. Dictionary of Modern Yoruba. London: University of London Press, 1958. Fanon, Frantz. Black Skin, White Masks. New York: Grove Press, 1967.

_____. The Wretched of the Earth. London: Penguin Books, 1978.

Hallen, Barry. “Analytic Philosophy and Traditional Thought: A Critique of Robin Horton.” African Philosophy: A Classical Approach. Eds. P. English and K. M. Kalumba. Englewood Cliffs, N.J.: Prentice-Hall, 1996. 216-228

_____. “Moral epistemology: When Propositions Come Out of Mouths.” International Philosophical Quarterly 38.2 (1998): 187-204.

_____. The Good, The Bad, and the Beautiful: Discourse About Values in Yoruba Culture. Bloomington and Indianapolis: Indiana UP, 2000.

Hallen, Barry and J. Olubi Sodipo. Knowledge, Belief, and Witchcraft: Analytic Experiments in African Philosophy. Stanford: Stanford UP, 1997.

Quine, W.V.O. Word and Object. Cambridge, Mass: MIT Press, 1960.
Wiredu, Kwasi. Philosophy and an African Culture. Cambridge and New York: Cambridge UP, 1980.


  1. W.V.O. Quine, Word and Object (Cambridge, Mass: MIT Press, 1960).
  2. Frantz Fanon, Black Skin, White Masks (New York: Grove Press, 1967) and The Wretched of the Earth (London: Penguin Books, 1978); Barry Hallen, “Analytic Philosophy and Traditional Thought: A Critique of Robin Horton,” African Philosophy: A Classical Approach, eds. P. English and K. M. Kalumba (Englewood Cliffs, N.J.: Prentice-Hall, 1996) 216-228 and The Good, The Bad, and the Beautiful: Discourse About Values in Yoruba Culture (Bloomington and Indianapolis: Indiana UP, 2000).
  3. Kwasi Wiredu, Philosophy and an African Culture (Cambridge and New York: Cambridge UP, 1980).
  4. Barry Hallen, “Moral epistemology: When Propositions Come Out of Mouths,” International Philosophical Quarterly 38.2 (1998): 187-204.
  5. The Dictionary of Modern Yoruba compiled by R.C. Abraham (1958) usually serves as the standard reference for Yoruba-English translations of this variety. Abraham treats “ooto” as a straightforward equivalent of the English-language “truth,” and the same is the case with “igbagbo”/“gbagbo” (233) and the English-language “belief”/“believe.” Both are examples of the understandably ‘loose’ translation equivalences that are a necessary evil for the conventional, cross-cultural translation of everyday equivalences that are a necessary evil for the convention, cross-cultural translation of everyday matters, and which cannot afford to take account of all semantic differences, even if they happen to be more than nuances.
  6. One expression used regularly for testing was danwo (‘try to do’).
  7. Since it may now be said that the various disputants are reconciled.
  8. According to Abraham (1958), “nwadi” is a participial conflation of the verb for ‘looking’ or ‘seeking’ with the noun “idi” (272) for ‘bottom,’ ‘base,’ ‘reason,’ or ‘cause.’ See also Barry Hallen and J. Olubi Sodipo, Knowledge, Belief, and Witchcraft: Analytic Experiments in African Philosophy (Stanford: Stanford UP, 1997) 70.

Barry Hallen

Barry Hallen is Professor of Philosophy and Chair of the Department of Philosophy & Religion at Morehouse College, Atlanta, Georgia. He is also a Fellow of the W.E.B. Du Bois Institute at Harvard University. His latest book is A Short History ofAfrican Philosophy (Indiana UP, 2002).