[ Pobierz całość w formacie PDF ]
LANGUAGE
&
COMMUNICATION
Language & Communication 27 (2007) 81–107
www.elsevier.com/locate/langcom
Linguistic relativism: Logic, grammar, and
arithmetic in cultural comparison
Christian Greiffenhagen
*
, Wes Sharrock
Department of Sociology, Roscoe Building, University of Manchester, Manchester M13 9PL, United Kingdom
Abstract
Linguistic relativism is the thesis that the grammatical structures of different languages imply dif-
ferent conceptions of reality. In this paper we critically discuss one form of linguistic relativism,
which argues that grammatical differences between the English and Yoruba language exhibit differ-
ences in how English and Yoruba speakers ‘see’ reality (namely in terms of ‘spatiotemporal partic-
ulars’ and ‘sortal particulars’, respectively).
We challenge the idea that linguistic relativism is an empirical thesis, i.e., a thesis that is substan-
tiated through anthropological examples. We show that linguistic relativism is based on two assump-
tions: firstly, that the purpose of language is to describe the world; secondly, that being able to speak
presupposes an ontological theory of the ultimate constituents of the world. We argue that the
attempt to extract the outline of that theory from the language inevitably distorts the portrayal of
language-using practice itself.
2006 Elsevier Ltd. All rights reserved.
Keywords: Cultural relativism; Philosophy of mind; Whorf; Wittgenstein
1. Introduction
This article contributes to a critical analysis of some empirical examples that are given
in support of cultural relativism, which we take to be the thesis that logic, grammar, and
arithmetic are, in some sense, relative or only locally valid. Our interest lies in showing
how hard it is to pin down just what ‘in some sense’ amounts to. In this paper, we will
*
Corresponding author.
E-mail address:
(C. Greiffenhagen).
0271-5309/$ - see front matter 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.langcom.2006.05.001
82
C. Greiffenhagen, W. Sharrock / Language & Communication 27 (2007) 81–107
focus on language and linguistic relativism.
1
As will be seen, however, this discussion mod-
ulates into a discussion of arithmetic, since the arguments about the relativity of language
are used to motivate a case about the relativity of number.
The kind of linguistic relativism considered here involves the assumption that different
grammatical structures and vocabularies imply different conceptual schemes or metaphys-
ics, which postulate the ultimate constituents of the material world. For example, accord-
ing to Whorf:
[...] the Hopi language and culture conceals a metaphysics, such as our so called
naive view of space and time does, or as the relativity theory does; yet it is a different
metaphysics from either. (
Whorf, 1956, p. 58
)
In other words, linguistic relativism is less interested in language per se than in investi-
gating which ultimate constituents of the world are identified ‘in’ the language. The rela-
tivist’s engagement with language is metaphysically motivated, based on the Whorfian
assumption that each language accommodates a distinctive, unified metaphysics. This
licenses the supposedly startling conclusion that inhabitants of different language commu-
nities inhabit worlds consisting of different kinds of things:
When [Whorf] uses [...] phrases [such as] ‘dissect nature’ and ‘the categories and
types that we isolate from the world of phenomena’, he is not thinking of fish and
fowl or flora and fauna; he is thinking, rather, of the metaphysical categories: thing,
event, relation, substance, quality, action, past, present, future, and so on. These are
the categories into which languages ‘dissect nature’, and Whorf’s relativism is the
claim that languages differ in the way in which they deploy such categories. (
Cook,
1978b, p. 5
)
We take it as plain that there are differences between languages, e.g., in their grammat-
ical structures or vocabularies. However, we question whether this necessarily implies var-
iable metaphysical commitments on the part of speakers of different languages. The reason
for this is not because we think that the metaphysics are uniform across languages, but
because we wonder (following Wittgenstein and Cook) whether speaking a language
depends on a metaphysics in the first place. In other words, the question for us is not
whether there are differences between languages (of course there are), but what this signi-
fies, especially whether epistemological or ontological implications comparisons can be
made of the basis of them – without, that is, the importation of questionable (at least con-
testable) philosophical assumptions to license reading these metaphysical implications into
the materials.
2
Thus we agree with Cook who notes:
It is, of course, an empirical fact that spoken languages do differ greatly in grammat-
ical structure, but it is an a priori philosophical claim that one can read off a meta-
physics from grammatical structure. (
Cook, 1978a, p. 13
)
As one of us once put it: relativists ‘‘are the kind of people who would attribute a geocentric theory of the
universe to us on the strength of the remark that we intend to get up tomorrow morning to watch the sun rise’’
(
Sharrock and Anderson, 1982, p. 111
).
1
In companion papers we discuss mathematical relativism (
Greiffenhagen and Sharrock, 2006
) and logical
relativism (
Greiffenhagen and Sharrock, unpublished
).
2
C. Greiffenhagen, W. Sharrock / Language & Communication 27 (2007) 81–107
83
In our view, the project of linguistic relativism only works by making several (unwar-
ranted) assumptions about language. Firstly, it assumes that the (only) function or pur-
pose of language is description. In other words, it assumes that language is a tool for
naming the things in the world. This assumption has an important corollary: if the purpose
of every language is description, then it is possible to compare whole languages of different
cultures (as they share the same ‘purpose’ of naming things in the world).
Secondly, in a typical intellectualist manner, linguistic relativism stipulates that all
description presupposes an ontological theory or a metaphysics – what
Whorf (1956, p.
221)
calls an ‘‘unformulated and naive world view’’ and what
Quine (1960, p. 22)
calls
‘‘the archaic and unconscious hypothesis of ordinary physical objects’’. In the form that
we encounter it here this results in the primitive idea that the relationship between lan-
guage and reality is effected by naming, and that the metaphysics of a language is mani-
fested in the kinds of things that its names stand for.
In contrast, we see language as inevitably tied to a people’s culture, i.e., to the world
that people live in and the practices that they engage in. Consequently, it is not possible
to compare languages per se but only aspects of languages. When looking at the empirical
examples offered by relativism, we find that the linguist either compares things that are
ostensibly not the same (e.g., science and witchcraft)
3
or that the compared things do
not seem to be significantly different after all (without stipulating a metaphysics).
In this article, we will try to exhibit that linguistic relativism is not an empirical but a
philosophical project – and that it is possible to give an account of the examples offered by
relativists without making the assumptions that the relativist makes. We will focus on one
central case of linguistic-cum-mathematical relativism, namely Helen Watson’s
4
initial
attempts (e.g.,
Watson, 1987, 1990; Verran, 2000a,b
) to contrast Yoruba and English
numbering. However, Watson is just one of many examples in the social studies of science
literature that are based on the idea that language is description and that language embod-
ies a theory of some kind (e.g.,
Bloor, 1976
).
5
2. Simplifying in order to create stark binaries
Watson aims to exposit the different kinds of things encoded in the grammar of English
and Yoruba. Her starting point is the assertion that
[...] predication leads language users to refer in particular ways and thus determines
what kinds of objects their language defines as constituting the universe. I show that
3
This was already observed by
Graves (1960)
in his critique of an early form of linguistic relativism expressed
by
Lee (1960)
: ‘‘When discussing differences between the sound-complexes applied in various countries to the
same reality, one should first make sure that it is the same reality. The words Brot, pain, and Pan are given only
one meaning in an English polyglot dictionary: namely, ‘bread’. But heavy German Brot, and light French pain,
and hard Spanish pan are not all the same reality’’ (p. 155).
4
Helen Watson is now Helen Verran. We will throughout refer to her as ‘Watson’, citing, where necessary, her
later papers as ‘Verran’.
5
It is also reflected in the deep rooted and general sociological suspicion of ‘ordinary language’ and in much of
the long tradition of ideology critique. Many of the tendencies are crystallised in contemporary ‘critical discourse
analysis’. In these approaches ‘ordinary language’ is seen as embodying a theory, but an at best na¨ve, or, most
often, a wrong one. The relativists share the assumption about the theory-dependence of language but are apt to
think that all the theories are ‘right’.
84
C. Greiffenhagen, W. Sharrock / Language & Communication 27 (2007) 81–107
English and Yoruba (a West African language) predicate in different ways, and con-
sequently postulate different kinds of objects. (
Watson, 1990, p. 283
)
Note that Watson is talking here about the whole of a language and not some specific
aspects, e.g., names of snow, cooking utensils, or wedding ceremonies. Furthermore, Wat-
son is comparing exactly two languages, English and Yoruba (rather than eight languages,
which might imply that she would have to specify how the languages are different and sim-
ilar). Watson thus employs a typical relativistic strategy which we might call ‘simplifying in
order to create stark binaries’. Through creating gross synoptic characterisations of
aspects of exactly two different languages, it is possible for Watson to make a simple
and stark contrast. For example, she claims that:
Talking about the world necessarily involves postulating the type of constituents of
the material world. (p. 288)
This stipulation is not taken as a starting point for a methodic exploration of how
many different ‘types of constituents’ each language (English and Yoruba) postulates
(perhaps three, five, or eighty-seven). No, it is further stipulated that each language will
(with respect to numbers) postulate exactly one type of constituent. In other words, by
limiting the possible choice of types of things to exactly one per language, Watson can
reduce variations and align them along a single axis. This allows her to portray the two
languages as standing in a simple contrast with respect to the types of things they
postulate:
The difference in meaning [...] is a difference in the type of material object which is
being postulated through use of a referring category. (p. 289; emphases added)
By reducing the differences between languages to one variable, all other differences are
obscured. It is this simplification that enables the transition from a very few examined
instances to an account of reference for a whole language that leads relativists as Watson
to think that encountered differences have to be consistently contrasting ontological ones.
If the analyst took greater care to look at the similarities and differences between lan-
guages, as well as the different kinds of differences within and between languages, the rel-
ativistic project would be much harder to get started.
3. The ontological theory ‘encoded’ in language
Linguistic relativism proceeds by stipulating a separation between words and concepts
associated with words. The latter is supposed to refer to our ideas or theories about what
makes the thing designated by the word what it is. For example,
Watson (1990, p. 283)
asserts that ‘‘how we count depends on how we conceive objects to exist’’. In other words,
she assumes that our capacity to learn and use language requires possession of an addi-
tional theory of the world, namely a theory that specifies what kinds of things our words
– and numbers – refer to.
How can the analyst identify this underlying theory? For the relativist the answer lies in
grammar. It is a further fundamental assumption of linguistic relativism that the grammar
of a language reflects the form of the underlying concepts of a language. Hence a charac-
terisation of the grammatical forms of a language should allow the anthropologist to make
inferences about the concepts that language-users have of their world:
C. Greiffenhagen, W. Sharrock / Language & Communication 27 (2007) 81–107
85
Whorf, like many philosophers, has the idea that one can read off a metaphysics
from the grammar of a language and thereby discern something about the thinking
of those who speak the language. (
Cook, 1978b, p. 7
)
Whorf thinks of a metaphysics as something that is contained in a language as an
inevitable feature of its grammatical structure. [...] he believes that grammatical cat-
egories lend themselves to a semantic interpretation employing metaphysical terms,
such as thing, event, process, relation, past, present, future, and so on, so that to
speak in a certain language just is to operate with, to think with, certain metaphys-
ical categories. (
Cook, 1978a, p. 12
)
Just as we are not denying that there are differences between languages, so we are not
denying that people have concepts and that these concepts may differ between different
cultures (and differ in different ways). However, following
Cook’s (1978a,b)
critique of
Whorfian relativism, we challenge the idea of language as a conceptual scheme, i.e., the
view that language embodies theories of reality. We are not denying that we have concepts,
but we are claiming that these concepts arise within language and do not underpin it.
6
For the relativist, the Whorfian assumption that the grammar of a language embodies
metaphysics is the bridge from language to language users. In other words, it is the link
from linguistics to psychology, sociology, and philosophy – from the structure of language
to the ways in which people think (especially the way they think of reality). Anthropolo-
gists such as Watson are not interested in only exhibiting certain grammatical differences
between languages (say, between the form and placement of otherwise equivalent number-
words in sentences), but want to demonstrate that what these numbers mean (i.e., what
species of material entity they are referring to) for language users is different.
For the relativist, in order to be able to speak ‘about’ the world one has to be able to
classify over the things ‘in’ the world. This in turn requires the possession of a theory ‘of’
the world (which specifies the criteria for the demarcation of the categories):
Learning to mean with language starts with classifying, but classifying does not
amount to making meaningful sentences. Meaningful sentences are uttered within
the framework of a particular theory of the world, a theory which the language used
in making meaning necessarily and inevitably encodes. (
Watson, 1990, p. 287
)
In other words, for Watson it is necessary to have a theory of the world as a precondi-
tion of being able to acquire language. This theory is encoded in (the grammar of) lan-
guage. Furthermore, this theory has the form of a proto-language:
Predication develops from the more complex of the classifying activities in what we
might call ‘proto-language use’ – classifying over actions of bodies. (p. 287)
Thus, for the relativist, in learning language the child is moving from learning to clas-
sify things to learning to make meaningful statements. The child therefore must have made
6
Neither are we denying that words have origins (etymology). However, we are questioning the role of that
origin in current language-use. As
Lenneberg (1953, pp. 465–466)
observes: ‘‘There is a metaphorical element in
language per se. The literal meaning of many metaphors, especially the most frequent ones, never penetrates
consciousness, e.g., everybody, in the face of, beforehand, breakfast, inside, already. The translation method,
however, distorts the significance of such forms of speech and often induces investigators to draw rather ludicrous
conclusions’’. In short, the meaning of a word is fixed by its current use, not its origin.
[ Pobierz całość w formacie PDF ]