A concept is that which is understood with a term (i.e., its “meaning”).1 All concepts are FR-concepts (see chapter 4). To grasp a concept involves being able to use it appropriately in different language games and to exploit its family resemblance to other concepts. We consider that a concept covers Frege’s sense and reference, as well as connotation (as this notion is used in literary studies),2 as well as meaning as use. Even if a philosopher is opposed to the use of concepts for the reason that it would inappropriately regiment or limit thinking, her or his interpreters and commentators still employ concepts to explain her or his philosophy.
Conceptual schemes (henceforth: CS’s) are frameworks or taxonomies of concepts.3 To know the meaning of a concept, one has to know the meaning of many concepts and many CS’s. Hence, the meaning of a concept is dependent on the meaning of other concepts and a concept is already a CS (holism). Hence, often the notions of concept and of conceptual scheme can be used interchangeably.
To discuss a specific subject matter one uses CS’s that elucidate that subject matter. A particular conceptual scheme is always tied to a particular language segment (which consists of a cluster of language games and is similar to Lyotard’s notion of a phrase regimen). Although in principle the meaning of a word/concept depends on all (possible) language games in which it might occur, in practice a critical mass will suffice.4 The composition of the critical mass will depend on the broader aims of the interpreter.
In the Western traditions, the notions of concept and conceptual scheme go back to the time of Plato and Aristotle.5 The notion of a primitive or closed concept is associated with the notion of ideal language. In chapter 4, we have defended that all concepts are open concepts, unless stipulated to be closed with respect to a particular application (for example, a formal language used in particular computer software). Open concepts are FR-concepts. They need not share a common property or essence; they have no precise definition; no strict borders; and they are intensionally incomplete, that is, adaptable according to their role in activities and theories. A conceptual scheme shares the features of concepts just listed.
Conceptual schemes at a high level of abstraction are referred to as meta-conceptual-schemes (henceforth: meta-CS’s). The difference between concept and meta-concept is relative. Relative to green and turquoise, color is a meta-concept. Relative to the concept color, appearance is a meta-concept. Most concepts are sortals; that is, they have (vague) identity conditions. Meta-concepts such as “thing” or “object” don’t have identity conditions unless further specified. One cannot count “things”; one can only count a particular kind of things.6
Meta-CS’s vary across traditions. A case study of the meta-CS’s “thing” and wu 物 is discussed later in this section.7 We consider that philosophical categories are meta-CS’s. Kant’s system of categories can be considered to be a proposal for one universal meta-CS. The notions of de-essentialization (family resemblance) and the problem of complete description apply to concepts, conceptual schemes, meta-CS’s, and categories. The choice and justification of a particular conceptual scheme is ultimately grounded in natural language (games) embedded in (de-essentialized) forms of life.
Philosophers easily tend to conceive a conceptual scheme as part of an ideal language. For example, both Quine and the Later Mohists attempt to provide strict definitions for the terms used in a conceptual scheme. But a conceptual scheme does not have to aspire to becoming a highly regimented (ideal) language. Providing an explanation of the interrelations of a number of central characters used in the Daodejing, together with some remarks on their relations to the scholarly Chinese language of its time, also presents a (sketch of a) conceptual scheme. Constructing a formalized CS that resembles an ideal language as defined by Frege may have important theoretical or practical significance. However, no matter how sophisticated the formal CS is, one cannot but use a natural language to judge its significance in relation to alternative proposals.
Two concepts are never the same, but they may be related by FR-judgments, as judged from a third (meta-)CS. There is an unavoidable infinite regress here. When a CS is judged from the perspective of yet another CS, the latter judgment can again be judged in terms of another CS. Whether the similarity between two concepts or conceptual schemes is superficial or more substantial is open for discussion. For example, how to establish whether two philosophers or traditions are using a similar concept of human rights.8 Then one needs a third CS to compare the two views. Hence, whether two concepts are very similar or not at all is a judgment that cannot be made independently of particular meta-CS’s. Idealizing CS’s may be helpful in clarifying what one is saying. It is similar for clarifying the meta-CS’s of two “factions,” but again one needs a third meta-CS to carry out the comparison. There is no ideal tradition-independent metalanguage that could allow a final judgment.
The interpreter may select a (meta-)CS from her or his own tradition to explain a particular (meta-)CS in another tradition. This is a common practice in intercultural philosophy. The family resemblance of (meta-)CS’s is grounded in mutually recognizable human practices. There is not a single overarching CS for all traditions (universalism), nor does one tradition (or individual) correspond to only one CS (relativism). It is commonly seen that scholars associate with a philosopher or a tradition no more than one conceptual scheme (either claimed to be universal or unique for that tradition). In the light of our plea for de-essentialization, it is crucial to realize that a large number of conceptual schemes are simultaneously at play in the thinking of a philosopher or in a tradition. Every philosopher or tradition is engaged with an indefinite manifold of CS’s, which cannot be integrated into one overarching architectonic CS.9 It is not possible to simply “add up” these schemes, because some will be mutually inconsistent (in terms of a third scheme in the cluster).
Furthermore, the ways in which a conceptual scheme receives determination are indefinite. Every scheme, no matter how well formalized or empirically adequate, is nothing more than a sophisticated abstraction from numerous language games of a tradition or a philosopher, the manifold of which can never be completely described. One can say that each CS fits a subject matter (or “a world”); nevertheless, each utterance about such a world can only be a co-production of many CS’s.
“On the Very Idea of a Conceptual Scheme” is the title of a highly influential article of Davidson (2001: 183–198 [1974]), in which he argued that conceptual relativism is incoherent, because the idea of different conceptual schemes makes no sense. Hence, Davidson concludes that the notion of conceptual scheme is useless, and that conceptual relativism and its related notion of incommensurability are indefensible. It is important to stress that our use of the phrase conceptual scheme is not subject to Davidson’s criticism of “the very idea of a conceptual scheme.” Davidson considers a CS as a conglomerate of true sentences representing all beliefs of the holder (person or tradition). In what follows we present his views, substituting conceptual scheme with web of belief (Quine’s phrase).10 The main differences between the two notions are:
1.Webs of beliefs (CS’s in Davidson’s sense) focus on (true) sentences. CS’s (in our sense) focus on patterns of (named) concepts. On this point we follow Graham (1992a).
2.Davidson’s notion of CS presupposes an ideal language into which a natural language is to be modeled. In contrast our notion of CS has appropriated the family resemblance theory of meaning.
3.We assume that one person or tradition is using an indeterminate number of conceptual schemes, which cannot be integrated into one scheme. In contrast Davidson lumps everything together into one all-encompassing web of (true) beliefs of a tradition or person.
Given Davidson’s assumptions, the idea of completely different (incommensurable) webs of beliefs is incoherent indeed.11 Davidson’s argument runs as follows. One might think that one could judge whether two schemes (webs of beliefs) were different by placing oneself completely outside all schemes. But that is impossible. To detach oneself from all schemes would be to detach oneself from language. Language is necessary for thought. Hence one cannot place oneself outside all schemes and look at them from afar. One might think that, if a language cannot be translated into one’s own, one is encountering a case of a completely different web of beliefs. But this is unimaginable. If we assume that the other language is totally untranslatable despite repeated efforts, how would one know it was a language in the first place (and not random noises)? How does one know that its “speakers” are humans? Failures of translation can only be made intelligible against a largely shared background.
Hacking (1983) has pointed out that, short of radical incommensurability, dissociation or divergence of different styles of reasoning does make sense. Divergence can be more eye-catching if there is a background of similarities. An example might be the divergences between the conceptual schemes or web of beliefs (both of them ordered and coherent) of the Aztecs and newly arrived Spaniards at the beginning of the seventeenth century, although there were quite a few similarities between the social organization of the Aztecs and the Spanish (Lockhart 1994). On the other hand, in case of “extreme” divergence, it will be difficult to construct a metalanguage in which the difference can be elucidated, and we may speak of incongruities in the sense of “out of place.” Talking past one another may be considered a modest case of divergence or incongruities that transcend differences.12
Though divergence can be large, there is never total incommensurability. How else could one pretend to explain Paracelsus’ ideas of treating syphilis or Averroës’ difficulties in translating Aristotle’s Poetica?13 Something like incommensurability makes sense, but only much later in the stages of interpretation. If there is a provisionally shared basis, one may note incommensurable CS’s, both locally and interculturally. For example, one may come to think that qing 情 in classical Chinese is, strictly speaking, incommensurable with the modern Western notion of emotion(s), but this can happen only after one has noted a lot of similarities between FR(qing 情) and FR-concepts in Western languages.
By proposing alternative ways of identifying webs of belief, numerous ways have been proposed to avoid Davidson’s conclusion that a completely different web of beliefs is not possible.14 However, we think that his basic intuition is correct, no matter whether one focuses on beliefs, sentences or concepts. This intuition is expressed in the following:
A creature that cannot in principle be understood in terms of our beliefs, values, and modes of communication is not a creature that may have thoughts radically different from our own; it is a creature without what we mean by “thoughts.” (Davidson 2004: 37)
A completely different web of beliefs or a completely different set of (meta-)CS’s makes no sense, because labeling it as such entails that one has already identified it as a web of beliefs or conceptual schemes and embedded it, however provisionally, in human practices, conceptualizations and other signifiers of the variety of human forms of life.
Wittgenstein once gave a reply to esoteric speculations concerning the possibility of a group of humans having an allegedly untranslatable language:
Let us imagine that the people in that country carried on the usual human activities and in the course of them employed, apparently, an articulate language. If we watch their activities, we find them intelligible, they seem “logical.” But when we try to learn their language, we find it impossible to do so. For there is no regular connection between what they say, the sounds they make, and their activities; but still these sounds are not superfluous, for if, for example, we gag one of the people, this has the same consequences as with us; without those sounds their actions fall into confusion—as I feel like putting it.
Are we to say that these people have a language: orders, reports, and so on?
There is not enough regularity for us to call it “language.” (PI §207)
We agree that what is described in the citation is not a language. Hence, it is not an untranslatable language. When there is enough regularity to call it language, the language and its CS’s are already partly (minimally) understood. There can be no such thing as a completely untranslatable language.15
Kuhn is primarily known for his Structure of Scientific Revolutions of 1962 and key terms such as paradigm and incommensurability. Here we consider his views about three decades later (Kuhn 2000). According to the later Kuhn, incommensurability is a necessary consequence of the nature of language and of cognition.16 It can be defined as differences in taxonomies of natural kinds. An account of this natural-kind-incommensurability should be part of a theory concerning the dynamics of scientific knowledge. He claims that this was part of his earlier views as well. And we add: such an account must be part of a theory of interpretation across traditions as well.
Incommensurability does not pose a threat to rationalism, according to Kuhn.17 Such a belief is very much a philosophical artifact and has little to do with the practice of science or interpretation.18 In practice, the same researcher or interpreter can work with incommensurable theories. In contrast with the standard reading of the early Kuhn, the later Kuhn does not assume that the whole lexicon is infected by the (very) different meaning of some central high-level abstract terms.19 Everyday matters can be described in an independent observation language that is easily translatable across traditions. For example, the Khoikhoi (or Khoekhoe, South Africa) are very good at finding their way because they “know the language of the stars,” which, they say, is not surprising because the seven sisters (the seven daughters of Tsiqwa, the Creator) are the star mothers from which the entire human race (i.e., the Khoikhoi more specifically) has descended. The Khoikhoi pick out the same objects as other people who use the stars to find their way.
Kuhn’s approach is primarily that of a historian of science and it is very similar to an approach with regard to the interpretation of ancient texts of whatever kind. To understand a sample of scientific knowledge or a text from the past, the historian has to learn another lexicon that systematically differs in some places from modern (scientific, philosophical) terminology. That this is needed becomes apparent when one reads old texts with modern eyes (either in a translation or in the original). A lot of it is rather easy to understand. For example, when reading chapter 11 of the Daodejing, one has sufficient agreement with the writer or composer of the text on the appearance of wheels with spokes or on the way one molds clay to make a vessel, and the like. However, there are phrases that are not accessible by simple FR-extension.20 For example, it is not easy to reach agreement on the translation of the last verse of chapter 11: wu zhi yi wei yong 無之以為用?21 However, according to Kuhn, one can learn expressions from an old lexicon, even if no suitable translation can be found.
Kuhn stresses that for a number of words, in order to understand them, one has to learn a cluster of concepts (an extended “critical mass” conceptual scheme), which are FR-connected in a way that is different across traditions. An example from Kuhn’s own research is that one cannot have an inkling of Aristotle’s understanding of movement if one does not also have some idea of his understanding of vacuum, matter, space, and related Greek words. Kuhn also gives the following everyday example:22
[The French word doux/douce] can be applied, inter alia, to honey (“sweet”), to wool (“soft”), to underseasoned soup (“bland”), to a memory (“tender”), or to a slope or a wind (“gentle”). (48)
When translating doux/douce, the translator’s choice highlights some aspects of the French intension at the expense of others.
Though the various translations offered above preserve truth value in appropriate contexts, none of them is in any context intensionally precise. (ibid.)
In learning new classifications of kind terms, one also learns to have a number of expectations concerning the behavior of these kinds. These expectations may differ somewhat from individual to individual and explain why in new situations different interpreters (including scientists, interpreters, and philosophers) may extrapolate what they have learned differently.23 This insight is also applicable to interpretation across traditions in the context of intercultural philosophy.
Putnam’s pragmatic realism involves conceptual relativity,24 which is very different from cultural or conceptual relativism (1987; 2004: 33–51). Different and even mutually inconsistent conceptual schemes can simultaneously play a role in communicative interaction, even in the same context for the same person.
Putnam asks his readers to consider a world with three individuals, x1, x2, and x3. How many objects are there in this world? In World I (“Carnap’s world”), there are three objects; in World II (“the Polish logician’s world”), there are seven objects, because in the latter World or CS, for every two particulars there is an object that is their mereological sum (1987: 19; 2004: 39). For the metaphysical realist, only one of these worlds can be the real one. According to Putnam’s pragmatic realism, questions of existence only arise after a CS has been chosen. According to Putnam (and Carnap 1950), it is up to us to choose between the CS of Carnap and that of the Polish logician.
The entities postulated by a scheme exist objectively,25 but they are not independent of conceptualization. This does not mean that the choice of a scheme “causes” the emergence of entities. But one can never say that something exists without already presupposing a particular ordering of the world (that is, without already having a particular scheme in place). The question “How many objects are there in this box?” would be meaningless if one does not specify a scheme that tells one how to count objects. A particular object can only exist provided that a classification or sorting has been assumed.26
According to Nelson Goodman, there is no principled difference between making a work of art and making a scientific theory, or, for that matter, making a translation or interpretation of an ancient text, making new concepts, making technological artifacts, making conventions, and making facts.27 Goodman’s views are rather similar to Putnam’s notion of conceptual relativity, except that he employs a more provocative terminology involving the word world.
The question whether the earth is moving or not is one of Goodman’s favorite examples (1987: 2–3). There seem to be two conflicting versions.28 There is the scientific version according to which the earth moves (around the sun) and the common sense version according to which the earth is at rest (and the sun rises every day). Goodman’s first step is to say that both ways of thinking are acceptable: the apparent contradiction does not imply that one version has to be dispensed with. Such a view is not an anything goes relativism, says Goodman. Surely one cannot let the earth move in any way one likes. It is also wrong to say that “the earth is at rest” was true long ago and now “the earth is moving” is true. Both versions are true, but for different worlds. Which world is the right one does not depend on one being true and the other not, but on the context of use a person brings to a version (and a world).29 Furthermore, the choice between versions depends on rightness instead of truth.30 Hence, Goodman concludes: If there is a world, then there are many, because the conflicting versions (“the earth moves,” “the earth is at rest”) cannot be true in the same world. On the other hand, because a world is understood to be a totality, these many worlds do not fit together into one bigger world. Hence, there is no world.
This kind of reasoning satisfies few philosophers.31 They complain that Goodman does not tell us how to individuate worlds or world versions—to which no doubt Goodman would reply that giving a criterion of individuation is making another world (and version); something that is difficult and yet allows for different right (and wrong) world makings.
Goodman’s plural use of “worlds” should not be associated with the notion of possible worlds; all his worlds are actual worlds. People make versions, and right versions make worlds. The making of worlds is not only with hands, but also with minds, with embodied minds and intentional hands. A strict separation between version and world is not possible. A true version is true in some worlds; but in no world is a false version true. Goodman writes:
How can a version be wrong about a world it makes? We must obviously look for truth not in the relation of a version to something outside that it refers to but in characteristics of the version itself and its relations to other versions. … Making right world-versions—or making worlds—is harder than making chairs or planes, and failure is common, largely because all we have available is scrap material recycled from old and stubborn worlds. … To say that every right version is a world and to say that every right version has a world answering to it may be equally right even if they are at odds with each other. (1984: 37, 42, 41)
There is no world independent of a version or a scheme. The notion of correspondence between something that is described and something that transcends description (THE WORLD) is incomprehensible. Someone who would like to mention a version-independent feature of the/a world has to use language, and hence also some world version.
How does one live in or with these worlds? In daily practice one jumps from one version to the other. There is nothing special about the word version; one may also use words such as: story, discourse, theory, conceptual scheme, language game, form(s) of life, and so on. Within one version there is no problem of referring to the world. But if one talks about different versions one simply forgets the world(s) for a moment.
Hence, Goodman would give this advice to the intercultural philosopher: Don’t be afraid of CS’s being incommensurable. There is nothing special about it. We use them every day. One just jumps from one version to another. A problem only arises if one wants to combine all versions into one big version or an overarching CS.
Hansen’s “mass noun hypothesis” concerning classical Chinese philosophy has repeatedly been discussed. There is no canonical formulation of this hypothesis, which has led to much confusion. We try to avoid reference to the syntax and semantics of (Chinese or English) mass nouns, as this small part of Hansen’s exposition (1983: 30–32) has caused the confusions. We rename Hansen’s hypothesis and formulate his mereology-hypothesis as follows (in Hansen’s words):32
[Suggestion:] attributing part-whole assumptions rather than one-many in explaining many other features of Mohist and Confucian and Daoist theories of language. (2007: 489n16)
[Radical hypothesis:] Suppose [pre-Qin] Chinese philosophers assumed something more like an ontology of stuffs [rather than an ontology of individuals]. When they talk of 10,000 wuthings [wanwu 万物] they probably do not mean 10,000 countable objects, but 10,000 nameable stuff kinds. (1992: 48)
[Conclusion:] Chinese ontology, I suggest, is mereological. (1983: 31)
Formulated in this way, the hypothesis suggests two contrasting meta-CS’s: one based on part-whole relations, the other on one-many or instance-kind relations. Ascribing to wu 物 the (Western) meaning “stuff kind(s)” entails ascribing to Chinese scholars a meta-CS’s different from the countable object meta-CS’s dominant in Western philosophy. According to Hansen, the mereological hypothesis may explain why Chinese philosophers are nominalists, as seen from the Western perspective. Other related claims include the view that the world is regarded as “a collection of overlapping and interpenetrating stuffs or substances” (30) and the view that terms for kinds of things are singular terms (35). These related claims we do not discuss.33
Empirical evaluation of the mereology-hypothesis is difficult, because Chinese philosophers did not make explicit statements to this effect. Probably Hansen would agree that the thesis can only be tested by applying it to the corpus of pre-Qin Chinese philosophers and then judge whether it works to make sense of what they are up to (presumably in terms that also make sense to a modern Western reader).34 This is what he aims to do in chapters 3 to 5 of Hansen (1983). Surprisingly, critics always discuss the first part of chapter 2 of this book, where Hansen makes some (rather unclear) suggestions to the effect that the “mass stuff view can be explained by specific features of the logical structure of Chinese nouns” (32). Detailed critical studies of Hansen’s mass noun hypothesis all focus on the syntax and semantics of Chinese and English (mass) nouns, instead of on the mereology-hypothesis.35 Hansen complains that Graham and others have misunderstood his views by focusing on the grammar of nouns in classical Chinese.
I hypothesized that the semantics of Chinese nouns may be like those of [English] mass nouns. (1992: 48n)
My speculative explanation actually started from my observation that pre-Han Chinese common nouns were not grammatical mass nouns. (2007: 489n16)
Perhaps critics of Hansen may be excused, because he did write: “The syntax of Chinese nouns is strikingly similar to that of mass nouns in English” (1983: 32); and “the grammar explains an alternative mass-stuff picture of reality” (54). However, the critics cannot be excused for overlooking that the issue of mass nouns is only the initial phase of Hansen’s whole project. As Fraser (2007: 451n28) pointed out, “Hansen indicates at least three times that the main burden of justification of his hypotheses is to be borne by the interpretive arguments of chapters 3–5,” instead of the discussion of mass nouns, which is only the “first step” (Hansen 1983: 9). If Hansen’s so-called mass noun hypothesis (whatever it is) is false, the mereology-hypothesis might still be true and find support in Hansen’s interpretations of the philosophy of language of Mohists, Confucians, and Daoists.
Fraser (2007) argued that Hansen’s mass noun hypothesis can be seen as a language-to-thought argument. It draws conclusions concerning the interpretation of an ancient philosopher from features of the language of that philosopher; hence, it is an example of the Sapir-Whorf hypothesis. However, according to Fraser, it is very difficult to find support for a language-to-thought argument and to distinguish it from other causes such as creative responses to philosophical puzzles. Because of the high level of abstraction and the regimentation of the data in the light of the theory, it is not easy to find examples that falsify Hansen’s thesis.36 Moreover, the mereology-hypothesis itself is not a language-to-thought argument. Only the suggestion that Chinese scholars may have been led to this view because of properties of their language would be a variant of the Sapir-Whorf hypothesis.
However, Fraser agrees with Hansen that “early Chinese thinkers employ a mereological ontology.” The mereology-hypothesis is “highly plausible” (2007: 420), although he can only give limited evidence.37 Perhaps the most convincing example he cites is the Later Mohists’ Canon B12, part of which reads:38
shu niu shu ma, ze niuma er. shu niuma, ze niuma yi.
數牛數馬, 則牛馬二。 數牛馬, 則牛馬一。
If we count oxen and count horses, then oxen and horses are two. If we count oxen-and-horses, then oxen-and-horses are one.
This looks like mereology indeed, but in his more detailed discussion of B12, Graham does not draw such far-reaching conclusions (1978: 362–364). Moreover, there is evidence against the mereology-hypothesis, as Fraser (440–442) points out as well. Pre-Qin scholars distinguish clearly between part-whole and instance-kind relations and appeal to similarities between individual particulars.
Harbsmeier, who has carried out extensive research concerning how nouns are used in classical Chinese, is rather critical of Hansen’s proposals. He writes:39
I find nothing in traditional Chinese literature that remotely suggests that the ancient Chinese ever thought of anything like Lesniewski’s mereology, or of apples and evening skies as one object scattered through time and space. In any case, Hansen does not provide any of the necessary detailed philological evidence to prove that the Chinese did think in this way. (1998: 312)
To Harbsmeier’s “detailed philological evidence,” Hansen’s reply is that Harbsmeier cites as evidence for his view texts in his question-begging translations (2007: 490n18) and “systematically confuses syntax and semantics” (1992: 48). Graham accepts Harbsmeier’s detailed survey, in which he distinguishes between count nouns, mass nouns, and generic nouns in classical Chinese at face value.40 He writes:
The mass-noun hypothesis is outdated, but something valuable survives: The tendency of Chinese thought is to divide down rather than to add up, to think in terms of whole and part rather than class and member. (1991: 276)
According to Hansen, the latter insight, which “survives the upheaval” (Graham 1989: 402), is “the whole of the mass-noun hypothesis” (1992: 48n).41
Putting aside the sinological discussion concerning the mass noun or mereology-hypothesis, we suggest that this case study illustrates the possibility of alternative meta-CS’s to present the structure of the world (Graham 1989: 402; 1991: 276):42
•On one side: reduction of cosmos and communities to aggregates of atoms and individual persons. This might be called the atomistic worldview.
•On the other side: variously divisible dao 道, li 理, and qi 气. This might be called the mereological worldview.43
Note that, in saying this, Graham is using a meta-meta-CS’s to talk about FR(thing) and FR(wu).44
Hansen’s mass noun hypothesis has met with severe criticisms. Cheng Chung-ying writes (2007: 558),
The assumption that ancient Chinese language is merely a mass-term language is gratuitous and groundless. All the ancient texts can be given subject-predicate or particular-general term interpretation, which is first-order predicate logic.
As suggested by Fung, Hansen may not have full knowledge of ancient Chinese and its grammar or its modern counterpart, otherwise he would not distort Chinese language to such a grotesque extent.
However, the fact that classical Chinese grammar can be given Cheng’s favored interpretation along the lines of first-order predicate logic does not entail that no other alternative interpretations are possible. Fung Yiu-ming follows Cheng, and says that Hansen’s interpretation is “startling”:45
It makes the ancient Chinese language committed to an ontology which has no physical objects and abstract entities; and this is quite awkward a view in the sense that it is contrary to the well-received view in the field of philosophy of language today which regards the concepts of physical object and abstract entity as a condition of “language-hood.” (2007: 527)
However, it is dubious whether current philosophy of language can be used as the final arbiter, if only because Hansen also appeals to it in justifying his views. The above criticisms of Hansen may be motivated more by a concern that his idiosyncratic thesis may present pre-Qin literati as having a less developed language. Cheng and Hansen both know that Quine suggested that a “mass noun phase” is a primordial stage in the development of language.
In brief, Quine’s speculations concerning the ontogenesis of reference is as follows. In the first stage, utterances such as “Milk!” “Dog!” “Cold!” “Raining!” are all on a par. There are no objects. There are only observation sentences in distinctive circumstances.46 In the second stage, utterances have a structure such as “Whenever there is swan, there is white swan.” The utterance states concomitant events, but there is no reification yet. Reification, the emergence of objects, only appears in the third stage: “Whenever there is (a) swan, it is white.” The introduction of the pronoun “it” is crucial in reaching the stage of reification.47 Hence Quine concludes,
The reification of bodies comes in stages in one’s acquisition of language, each successive stage being more clearly and emphatically an affirmation of existence. (1992: 7)
The very notion of an object at all, concrete or abstract, is a human contribution, a feature of our inherited apparatus for organizing the amorphous welter of neural input. (1992: 6)
I see all objects as theoretical. (1981: 20)
Quine’s view that “object” is a theoretical notion is not idiosyncratic. Other Western philosophers have made similar points, emphasizing the relativity of the notion object or thing to human interests. For example, according to Chomsky (1995: 30):48
What is a thing, and if so what thing it is, depends on specific configurations of human interests, intentions, goals and actions, an observation as old as Aristotle.
Even within one tradition, object is an FR-concept. Putnam writes:49
“Entity,” “object,” “event,” “situation,” “fact,” “property,” etc., have not one fixed use but an ever expanding family of uses.
This “ever expanding family of uses” has to be extended across traditions to include hypotheses concerning an FR-relation between thing and wu (and Hansen’s “stuff-kind”). Across traditions, object (and/or thing, event, process, kind of thing) and wu form a cluster of FR-concepts. The remarks of Quine, Chomsky, and Putnam would not be much different if we replace “object” (or “thing”) by wu in the citations.
The case study discussed in this subsection shows that the notions of object and wu might be embedded in quite different meta-CS’s. “Object” is a theoretical notion relative to a particular meta-CS’s that could be different; for example, a wu-meta-CS’s that “divides down” instead of “adding up.” We do not intend to give the last word on whether the mass noun or mereology-hypothesis better suits Chinese classical philosophy (thinking). The example merely serves to show what considering alternative meta-CS’s would be like. Other possible alternatives at this high level of abstraction might be alternatives to FR(experience) or FR(appearance).50 When interpreting across traditions, one has to be prepared to come across very different meta-CS’s. But the latter will have to be understood in terms of the interpreter’s (extended and modified) meta-CS’s.
Kuhn’s notion of lexicon,51 Putnam’s notion of conceptual scheme and Goodman’s notion of version can be regarded as extensions of our FR-notion of conceptual scheme. Wittgenstein’s notion of grammar also belongs to this cluster. Our notion of CS draws on the work of Kuhn, Putnam, and Goodman, but it has a number of distinctive features, which can be summarized as follows:
1.A limited conglomerate of related concepts is a conceptual scheme. How to delineate a CS is a matter of interpretation in the light of the broader aims of the investigation.52
2.People use indefinite manifolds of schemes simultaneously—manifolds that can neither be described nor formalized in their totality. There is no single overarching meta-CS.
3.Adherence to particular CS’s is ultimately grounded in the human form(s) of life and their language games.
4.To judge the significance of a particular CS or to compare two CS’s, another CS is needed (and so on).
5.Each right scheme must fit “a world,” but each utterance about such a world is a co-production of numerous schemes.
6.Speaking about similarities and differences always takes place against the background of numerous CS’s.
Our interpretation and extension of Wittgenstein’s notion of forms of life is a natural development of the views we hold concerning FR-concepts. In order to emphasize that there is nothing special about the words “form(s) of life,” we introduce this Wittgensteinian notion via the notion of lifeworld as introduced by Husserl and used by Habermas (and several other influential German philosophers). Furthermore, the following expressions also serve a similar purpose: always-already-being-in-the-world, background, common sense, episteme, the everyday, habitus, moral order, prejudices, structuration, third space, version cum world, world image/picture (Weltbild), world of practical realities.53 When taken in a de-essentialized way, all these phrases have some similarity to form(s) of life. Without wanting to be hegemonic, one may speak of form(s) of life and its congeners. Such congeners can also be found among some interpretations of Chinese concepts.54
The notion lifeworld (Lebenswelt) has a central place in Husserl’s later writings.55 He argues that science is grounded in the everyday lifeworld shared by all humans. The lifeworld is a complex sphere encompassing the purely perceptual world, the everyday world of tools and tasks, the world of goals, interests, customs, and community, as well as ideal/abstract entities insofar as they make up a supposedly unitary tradition.56 In contrast with the supposedly objective-scientific world and other theoretical constructions, the lifeworld is given, immediately and directly. It is the foundation or source of sense, the subscientific world of living that the sciences naively take for granted. The lifeworld is the realm of pretheoretical life out of which science is constructed by abstraction from everyday beliefs. It is open-ended unlike the self-referencing, closed abstractions and idealities of scientific (or philosophical) theories. Scientific validities and theoretical praxis flow back into the everyday lifeworld. Active accomplishments become sedimented, presupposed, and habitual. The lifeworld absorbs scientific truths and thus expands.
The lifeworld does not encompass the abstract world of bodies and events, the so-called objective nature, which actually results from idealization (in physics). Rather, it is the ground of all theoretical abstractions (including physics and mathematics). The lifeworld is an accomplishment of a community of persons recognizing each other as normal, as similar to oneself and to one another, and as in rapport. The things of the lifeworld, taken exactly in the way they present themselves to us in this world, are not anonymous objects as in the case of the natural sciences, mathematics, or chess. Rather, they are specters of intuitive experiences. The lifeworld is the most general world. It is the most general community of persons, and serves as the presupposed background of all the special, institutional worlds that may arise.
Husserl’s notion of lifeworld is introduced specifically to provide a habitus for science. Its family resemblance with the notion of manifest image is obvious. Habermas’ notion of lifeworld comes closer to Wittgenstein’s notion of form(s) of life (Habermas 1985, part VI), given the account of form(s) of life we present below.57 According to Habermas, the lifeworld is the culturally transmitted and linguistically organized stock of meaning samples. He considers communicative interaction and Schütz’s culturalized notion of Husserl’s lifeworld to be complimentary, to which he explicitly adds the aspects of solidarity and cultural change. Habermas’ lifeworld takes for granted the attitudes of common sense. This is to be understood in a radical sense: the lifeworld is self-evident and cannot become problematic. It is prior to any possible disagreement and cannot become controversial as intersubjectively shared knowledge can; it can only collapse and disintegrate. The members of a collective count themselves as belonging to the lifeworld in the first-person plural “we.” There is naive trust in the lifeworld. What can be thematized remains encompassed within the horizons of the lifeworld, however blurred the latter might be. The lifeworld forms the indirect context of what is said, discussed, and addressed in a situation. It is intuitively present, familiar, and transparent; a vast and incalculable web of presuppositions that have to be satisfied if an actual utterance is to be at all meaningful (valid or invalid).
Hence, the lifeworld always remains in the background; it is the unquestioned ground of everything and the unquestionable frame in which all tasks, problems, and questions are located. The lifeworld owes certainty to a social a priori built into the intersubjectivity of mutual understanding in language. Situations change, but the limits of the lifeworld cannot be transcended. The lifeworld forms the setting in which situational horizons shift, expand, or contract. It circumscribes action situations in the manner of a pre-understood context that is not addressed. It forms a context that is boundless, and draws boundaries. For participants, the lifeworld is a context that can neither be gotten behind, nor be exhausted. Every definition of a situation is an interpretation within the frame of what has already been interpreted, within a reality that is fundamentally and typically familiar.
Both Husserl and Habermas provided lengthy discussions about the lifeworld. In contrast, the expression Lebensform (form of life or life-form)58 occurs only a few times in Wittgenstein’s writings.59 However, there is a vast secondary literature that assigns a central role to this notion in Wittgenstein’s philosophy. In this subsection, we first present passages from Wittgenstein, then we elucidate these passages. In the next subsection, we add important modifications and extensions of Wittgenstein’s prenotional notion of form(s) of life.
Staying as close as possible to Wittgenstein’s own words, one might say: Form(s) of life are what is given, the certainty of firm regular forms or rules of action, which accompany (make possible) language (games).60
What has to be accepted, the given is—one might say—forms of life. (PPF §345)61
I would like to regard this certainty, not as something akin to hastiness or superficiality, but as (a) form of life. (OC §358)62
It is characteristic of our language that the foundation [Grund] on which it grows consists in firm forms of life [fester Lebensformen], regular ways of acting. Its function is determined above all by action, which it accompanies. (PO: 396f)63
In his early writings, Wittgenstein uses the expressions culture (Kultur) and form of life (Form des Leben) interchangeably.64
Imagine a use of language (a culture) in which. … We could also easily imagine a language (and that means again a form of life). …
In some cases, the emphasis falls on the relation of form of life and specific language games:65
The manifestations of hope are modifications of this complicated form of life. (PPF §1)66
This reaction [viz. “the pupil reacts to it thus and thus”], which is our guarantee of understanding, presupposes as a surrounding particular circumstances, particular forms of life and of language. (RFM VIII: §47)
Most of the occurrences of Lebensform stress the close relationship between language and forms of life:67
And to imagine a language means to imagine a form of life. (PI §19)
The speaking of language is part of an activity [Tätigkeit], or of a form of life (PI §23).68
What is true or false is what human beings say; and it is in their language that human beings agree. This is agreement not in opinions, but rather in form of life. (PI §241)69
That people understand one another and themselves is due to their common participation in certain patterns, modes, ways, shapes, or forms of life. Growing up is growing into form(s) of life.
A form of life is a way of living, a pattern of activities, actions, interactions and feelings which are inextricably interwoven with, and partly constituted by uses of language. … It includes shared natural and linguistic responses, broad agreement in definitions and in judgments, and corresponding behavior. (Baker and Hacker 2009: 74)
Hence one can think of form(s) of life as the whole of the moral, social, historical, communicative, mythical, and private discernments and orders that ground and create these orders—the always already being in these form(s) of life.70 Human being(s) and form(s) of life are interconnected expressions. Growing up in form(s) of life includes learning to see forms of life one does not participate in.
Form(s) of life refer(s) to the complex of natural and cultural circumstances, which are presupposed when using language and understanding the world. It is what makes meaning possible. “Words have meaning only in life” (Wittgenstein, MS 137: 41b). It is “the context in which signs receive their meaning and practices their justification” (Barry 1996: 119). It “provides both the context of meaningfulness and the standards of justification in terms of which anything at all could be said to be either right or wrong” (Hinman 1983: 339). It is a system or pattern of (linguistic and nonlinguistic) activities that provides the context of specifically living situations (Glock 2000: 69). However, there is no precise explanation of forms of life. A form of life is a given unjustified and unjustifiable pattern of human activity. It is “the ultimately unsystematizable complex of actual societal life on which any provisionally formulable regularities or rules of behavior are based” (Margolis 1987: 132).71
To these general acknowledged features of form(s) of life, we add the following further elucidations. All language-mediated inquiry is dependent on the tacit contingencies of forms of life. It is only on the basis of this and against the background of this that one is able to pursue the objectives of science and other theoretical praxes (cf. Husserl). Form(s) of life is/are an integral whole. Things like forms of experience (the subjective) and an independent physical or material domain (the objective) can be distinguished only as “theoretical” (and revisable) abstractions.
If one starts from certainties one can give reasons.72 But there is an end to giving reasons: the end is what is given in the form(s) of life; it is where my, your, his, her, one’s, our, their spade is turned.73 Questions (scientific or otherwise) can be raised about anything. But these questions cannot but be asked from within the certainties of form(s) of life. The certainties or secure attunements associated with particular form(s) of life have a similarity relation to Gadamer’s Vorurteil (prejudice) and Heidegger’s fore-conception (1927: §32). One could say that Heidegger’s fore-conception and Gadamer’s Vorurteil are part of the lifeworlds or forms of life a human being participates in. But they are not part of a fixed rock bottom. Strictly speaking, Vorurteile change in every interhuman encounter. Wittgenstein writes in On Certainty:
I distinguish between the movement of the waters on the river-bed and the shift of the bed itself; though there is no sharp division of the one from the other. (OC §97)
And the bank of the river consists partly of hard rock, subject to no alteration or only to an imperceptible one, partly of sand, which in one place and now in another gets washed away or deposited. (OC §99)
Wittgenstein’s forms of life have often been given a conservative association of defending the status quo or tradition. Although Wittgenstein is only thinking of epistemic speculation, not of a blueprint for social action, his metaphor of the “river-bed of thoughts” (OC §97) can be read as allowing radical acts (Hekman 1999: 438).
It cannot be sufficiently stressed that form(s) of life should never be understood as something static. Though form(s) of life may be referred to as bedrock, one should not understand forms of life as passive, as being given absolutely. Though bedrock points to secure attunements, in certain circumstances, it may give way (and similarly for rockbottom). Moreover, over time, as the metaphor entails, the rock bottom will change.74 But that does not make the secure attunements less secure when one’s spade is turned here (or there). Hence, though it is correct to say that certainties are grounded in form(s) of life, the latter is not to be thought of as fixed and static.
Perhaps the major disagreement among interpreters is the question whether form(s) of life should be thought of in the singular or the plural (both forms occur in Wittgenstein’s writings—see aforementioned citations); whether it should be thought of in terms of natural anthropology or cultural anthropology; or alternatively, in terms of an empirical notion or a transcendental concept. Our brief answer to such questions is: “Both!” This reply, together with our remark that a similar story could be told using other expressions (lifeworld, etc.) distinguishes our use of form(s) of life from other uses and closely relates it to the notion of family resemblance.75 By writing “form(s) of life” with the pluralizing “s” in brackets, we mean to refer to the singular and the plural at the same time, and to emphasize that boundaries between traditions or forms of life are porous. Every ordering is contingent, subject to the historical development of conceptual schemes, language games, and interacting life-forms. It is also meant to convey that the prenotional notion of form(s) of life should be taken to be empirical as well as transcendental grounding, to be moral and cognitive basis for everything else, to have universal as well as local range. There are both one and many human forms of life. It would be incorrect to speak of many human forms of life, because all have in common their humanness. There is one human form of life as distinct from the form of life of lions. It would also be incorrect to talk about one human form of life, because similarities and differences crop up and disappear without there being a common core.
One can imagine an innumerable number of (human) forms of life. But there are no criteria that could specify in advance what could delineate the form of all possible forms of life. Since forms of life are both the source and limit of meaning, one cannot step outside all of them to survey them. There is no position outside forms of life to judge a particular form of life, but this does not preclude the possibility of judging one form of life from the perspective of others. Moreover, one can make new language games and forms of life. Furthermore, the moment different forms of life come into contact, they start to interact and to change.
All human beings (all societies) are busy engaging in and dealing with many different forms of life. Even if a group of people does not have (did not have) obvious “neighbors,” one can still discern a variety of forms of life concerning castes, clans, sexes, genders, ancestors, spirits, animals, ghosts, transient states of existence, or such like. A particular division of (types of) forms of life has only local relevance, but being familiar with many forms is inherent to talking among and about humans.
The notion of form(s) of life should not be essentialized in the way that was or is common for the notion of culture. Forms of life are not related to and differentiated from one another as different genera or species might be; they are not discrete and separate entities, which tangentially impinge on one another without overlapping.76 There is a transcendental sense in which the common place question, “‘How to make sense or understand a culture other than one’s own,’ doesn’t make sense, because it uncritically presupposes that ‘understanding one’s own culture’ does make sense.” There is also the more empirical observation that “parts of ‘our’ culture may be quite alien to one of ‘us’; indeed some parts of it may be more alien than cultural manifestations which are geographically or historically remote.”77
Anything one says is from within form(s) of life. As Williams puts it:78 “the limits of our language mean the limits of our world.” Talking about form(s) of life would require assuming an external vantage point—and this is impossible. Therefore, form(s) of life can be seen as transcendental precondition or limit of human activities. Forms of life are what all human beings share in virtue of being members of the human life-form (aside from a shared biological nature). On this view, Wittgenstein’s investigation of rule-following is considered to be a transcendental inquiry providing insight into how people comport themselves.79 One can only explore the borders of the transcendental empirically using available concepts. This exploration may hint in the direction of where the border is, but direction is a metaphor that cannot be cashed out easily.80 No matter how extensive the empirical investigation, one cannot reach the border, let alone the other side. Any suggestion to the point that empirical research can determine why human form(s) of life are this way rather than that way misses the point.
In his discussion of these issues, Williams (152) refers to “transcendental facts.” However, talking about facts already makes the border and the other side of it more concrete than can be said. There is perhaps more truth in Sacks’ (1997) suggestion that, instead of some transcendental reality dictating the structure of any empirical form of experience, “such determining as there is seems to go in the opposite direction—from the empirical to the transcendental” (177). Instead of transcendental constraints, Sacks proposes the notion of transcendental features, which indicate local limitations on conceivability that do not translate into limitations on possibility. These transcendental features can themselves change over time and in a way that is not subject to a prior constraint.
Whether transcendental features that are contingent can be said to be genuinely transcendental is a matter of debate. It might be suggested that without transcendental constraints governing possible variety, there is no reason for trusting the possibility of significant common ground between different traditions. However, the problem here is not the reduction of the transcendental to the historical. The real danger is that what is contingent is easily thought of as truly transcendental.81 The question is not how to give content to talk about limits or transcendental facts in some general sense, but to show what can be said about it. Then it will become clear that the empirical (or natural) and the transcendental (or conceptual) features cannot be strictly separated.
Lear (1986: 293) has suggested that one should steer a middle course between the “(too) empirical (and probably false) [and the] transcendental and vacuous,” but he assumes (as Williams does) that logic provides an abstract formulation of rules which must be obeyed in actual cases if people’s actions are to be an expression of mindedness. He assumes that, though local practices might diverge, one could not encounter others who reject our logical principles (297).82 As we have argued in chapter 2, “standard” logical principles can be assumed in most contexts, but we may encounter situations in which these are not observed. Hence, in particular interpretative contexts, logic is also part of the hermeneutic circles or holisms of interpretation.
A variety of writers give, en passant, hints toward the problematic nature of drawing sharp boundaries between the empirical and the transcendental, or between the local and the universal. The suggestion that the empirical or natural becomes indistinguishable from the transcendental can be found in the work of Merleau-Ponty and Gadamer when they are criticizing Husserl.83 Wittgenstein expressed similar views. In Wittgenstein’s mature thought, the transcendental and the empirical appear to form a coherent whole.84
The limits of human life-form(s) are given by the possible family resemblances that crop up and disappear in the different practices or life-forms of human beings and their encounters with one another.85 What is similar should not be understood as some particular thing that is biologically, psychologically, or otherwise shared by all human beings. What is similar in human practices and conceptual schemes is what human beings would recognize as similar in interhuman contacts—a similarity that is, in a way, transcendentally grounded, but the content of this grounding remains tied to the local situation of (potentially actual) encounters between you’s and me’s. First contacts are a good example illustrating how the condition humaine is not a scientific fact. It is both an empirical and a transcendental precondition for being a human person that one knows the certainties of particular form(s) of life and that one is capable of recognizing and dealing with an indefinite variety of human behaviors and practices. Because one participates in forms of life, one can interact with other form(s) of life and interpret across traditions.
The understanding of conceptual schemes and form(s) of life as outlined in this chapter, in particular the conclusion that family resemblances of form(s) of life are the empirically supported transcendental condition of possibility for any interpretation, has important consequences for understanding the concept of similarity. First of all, when speaking about similarities and differences one has to be aware that such observations are always made in connection with numerous conceptual schemes, which are needed to express the similarities and differences. Secondly, there can be good reasons why a particular conceptual scheme is better than another, fits better among other schemes, which have already gone through a selection process of fitting. However, there is no conceptual-scheme-free or context-free ideal limit, allowing all “right” conceptual schemes to be eventually integrated into one “true” super-scheme.86 Thirdly, there can be no differences without similarities. This much follows from Davidson’s arguments (elaborated in chapter 10) and has been realized by other thinkers. Heidegger remarks: “In comparing, the two ‘things’ to be compared are somehow already equated with one another insofar as they are selected and presented as what is to be compared” (1944/45: 42). Heidegger’s discourse is: prior to comparing there must already be something in which the two are the same. Hence, “there always lies in the approach to the comparison a decision about what is the same” (43/138).87 In our discourse, a quasi-universal is the something in terms of which two things are compared. What this something is, is usually rather undetermined and evanescent. For example, if one aims to compare Aristotle’s arête and Confucian de 德, it has already been assumed that both have some similarity to virtue, understood as a form of moral excellence. The manifestations of virtue or arête or de 德 and their embedment in philosophical traditions and their conceptual schemes and forms of life may differ substantially. Nevertheless, the interpreter claims to have constructed a metalanguage in which the family resemblance of CS(virtue), CS(de 德), and CS(arête) makes sense.
The evanescent something in terms of which things are compared is, most generally speaking, that they are part of the mutually recognizable human practices or language games.88 Very different practices can be recognized as human practices because they show some similarity with practices the interpreter is familiar with. “We can see in different ways of life something like our own, so that we can find at least some sense in them” (Bolton 1982: 282). There will always be enough (perceived) similarities to make a start at interpreting differences (cf. first contacts). At the same time, one should always be ready to revise one’s secure attunements concerning this or that (assumed) quasi-universal. One might think that question and answer is a language game all humans are familiar with, but this may not be true for speakers of the Fore language (mentioned in chapter 2). Inductive theories may tell us that all human beings give attention to rites de passage (major events of life, for example death). However, even if the latter is true, one should be prepared to come across somebody like Zhuangzi who confronts death with exaltation.89
In a well-known passage Wittgenstein wrote:
Shared human behavior [gemeinsame menschliche Handlungsweise; earlier English translation: “common behavior of mankind”] is the system of reference [Bezugssystem] by means of which we interpret an unknown language. (PI §206)
There is an extensive secondary literature concerning the interpretation of this passage and the surrounding text.90 We follow those commentators who argue, on the basis of Wittgenstein’s corpus, not merely on the basis of this passage, that, as Hintikka and Hintikka put it (1986: 209): “There is in Wittgenstein’s writings no evidence that he believed in a way of behavior common to the entirety of mankind.” Therefore, contrary to what it may seem to suggest, the cited sentence does not imply that Wittgenstein presupposes the existence of some behavior common to all human beings and only human beings. It does not imply that certain patterns of linguistic behavior are realized in all languages or that some behavioral regularities prevail among all and only human beings (von Savigny 1991: 118). One piece of evidence in support of this is that, in one of Wittgenstein’s manuscripts, the passage cited is preceded by:
There is perhaps a people who possess no form of life corresponding to our “greeting” // who possess nothing corresponding to our “greeting.” (MS 165: 110)
Wittgenstein is not assuming that greeting is a practice all human forms of life share. The language game of greeting will often “work” as a quasi-universal, allowing for different manifestations “in the small,” for example, greeting by pressing noses. However, one can imagine coming across a community of humans for which one cannot identify a language game fitting an FR-extension of the language game of greeting.
There are serious problems with the translation of gemeinsame menschliche Handlungsweise as “common behavior of mankind” in the earlier editions of PI, a phrase repeated hundreds of times in the secondary literature. There are at least three different ways to interpret the German passage (cf. von Savigny 1991: 113):
•shared behavior (practices, ways of acting, forms of life) common to all humans;
•shared behavior (practices, ways of acting, forms of life) common to “the others” and the interpreter;
•shared behavior (practices, ways of acting, forms of life) common among the others.
The English translation “common behavior of mankind” suggests the first meaning, which, as von Savigny (1991) correctly points out, does not fit the context. The revised rendition “shared human behavior” retains ambiguity. Von Savigny opts for the third reading. We advocate something between the second and the third reading. Focusing only on the third reading leaves open the question how the interpreter has access to the others’ shared practices. For the point Wittgenstein wants to make, the third reading seems to fit best, but the second reading needs to be included in order to be able to interpret the others’ behavior at all. A further issue is a possible difference between Handlung (action) and Benehmen (behavior). The meaning of Handlungsweise is a way of acting. We prefer to translate it as practice. Further, in quite a few contexts, common is the correct translation of gemeinsame, but the latter can also mean shared or mutual. On the other hand, common can mean usual or familiar, which is not possible for gemeinsame. Instead of “shared human behavior,” we speak of “mutually recognizable human practices.”91
The passage just quoted is preceded by such a passage:
Suppose you came as an explorer into an unknown country with a language quite unknown to you. In what circumstances would you say that the people there gave orders, understood them, obeyed them, rebelled against them, and so on? (PI §206)
Perhaps on the first view one perceives nothing similar, but this cannot be the case for all the language games the interpreter is familiar with. Some of them must be FR-extendable to the unknown practices and CS’s. Absence of practices provisionally taken to be quasi-universals is possible, but this can only be observed against the background of a majority of family resemblances of practices and language games. It is the recognizable practices of humans as just elucidated that lie at the basis of similarities of (human) forms of life.
From the point of view of one language or one cluster of forms of life, the practices of human beings always show similarities (because they are human practices). Were this not so, communicative interaction or interpretation across traditions would be impossible. It is a necessary requirement for communication (or translation) that these similarities appear there. That there will always be relevant similarities between human forms of life is both an empirical and transcendental “fact.” However, what these similarities are is dependent on the forms of life compared and the conceptual resources available to those who make the comparison. Furthermore, as Goodman remarks:
Comparative judgments of similarity often require not merely selection of relevant properties [i.e., aspects of similarity] but a weighing of their relative importance, and variation in both relevance and importance can be rapid and enormous. … Circumstances alter similarities. (1972: 445)
The judgment of the relative importance of multiple aspects of similarity is made by a particular interpreter, who has his or her own background. This contributes to the underdetermination of interpretation.
Any understanding or account of what is similar across or within forms of life should be particularized; that is to say, should be limited to a few or only two (clusters of) forms of life. It should not be an attempt at grasping universals (in a Platonic, Kantian, hermeneutical, sociobiological, or whatever sense). What is similar has to be claimed again and again in all human interactions—where claimed should not be understood in individualistic terms, but as part of a complicated process of triangulation between at least two (groups of) human beings, their Umwelt, and their history and background (including more human beings).
The question might be raised how the proposition that human beings share similar responses to a diversity of forms of human life can be compatible with the suggestion that there is no core or essence to human behavior. They are compatible because similar does not mean the same. There are many similarities, but no universals, no cores. Moreover, different groups or individuals will observe other similarities, because these similarities are embedded in different conceptual schemes and forms of life. Similarly, in the practice of interpretation or comparison, the choices made by interpreters as to which FR-concepts are to be extended to quasi-universals differ. Theoretically, ontologically, there are no cores. However, on pragmatic grounds one can agree to use lhenxa as an Urfarbe,92 instead of green and yellow, or to use as a “standard” a traditional variant of zhengyi 正义 or ukulunga or geregtigheid instead of justice.93 Such pragmatic decisions are always subject to revision or contestation and will be understood differently in different languages.