But it is the second reason for Whorf's fall from prominence which is immediately pertinent to our present concerns. What was most startling about Whorfian relativity was that it was held to be true of particular human communities in the real world. A confusing characteristic of the chief philosophical discussions of radical translation, the inscrutability of reference, and the very idea of a conceptual scheme is that they are cast as thought-experiments insulated in principle from empirical verification and objection, a factor routinely ignored in anthropological borrowings from and disagreements with the philosophers. Whatever success the philosophers might or might not achieve, Whorf bluntly urged as a matter of fact that Hopi language does manifest basic features entailing that Hopi thought, for example in its treatment of time, diverges markedly from the language and thus thought employed by his own audience:

after long and careful study and analysis, the Hopi language is seen to contain no words, grammatical forms, constructions or expressions that refer directly to what we call 'time', or to past, present, or future, or to enduring or lasting...or that even refer to space in such a way as to exclude that element of extension or existence that we call 'time', and so by implication leave a residue that could be referred to as 'time'. Hence, the Hopi language contains no reference to 'time', either explicit or implicit.

The problem is that the best current work on Amerindian languages seriously undermines Whorf's results: both his data and their interpretation are, let us say, highly questionable. Most devastating, perhaps, is Ekkehart Malotki's eloquent juxtaposition (without further comment) of part of the passage from Whorf just quoted with this Hopi quotation from Malotki's own field notes of 1980: 'then indeed, the following day, quite early in the morning at the hour when people pray to the sun, around that time then he woke up the girl again'. Malotki's magnum opus (the very length of a book entitled Hopi Time (677 pages) speaks volumes) exposes Whorf's alternative 'thought world' as a complete fantasy: 'while many other contributing factors instrumental in disambiguating present and past time interpretations of Hopi nonfactive verbs cannot be detailed here, suffice it to say that Hopi speakers never consider themselves at a loss in determining whether a particular utterance refers to past, present, or future time'.

Thus much of the force of a supposedly 'hard' case dissipates: if anyone is inhabiting a distinct world, it isn't the Hopi, at least not for the reasons Whorf pretended he had unearthed. If the Hopi are profoundly alienated from main-stream American society, this may have a little more to do with economic and political oppression than with their delightful tense-system.

We have seen that Whorf himself was not at all averse to taking what he called 'culture', as opposed to 'language', into consideration in his study of the Hopi. In general terms, such catholicity is, of course, all to the good; but no consistent advocate of 'guidance and constraint' can afford to pay 'culture' the attention it would otherwise demand, if any progress is to be made towards charting the connections between language, thought and society. As a matter of historical fact, scholars reacting to the hypothesis of linguistic relativity were acutely aware of a set of problems closely resembling this dilemma. In the immediate aftermath of Whorf's celebrity, numerous linguists and anthropologists were rightly much exercised by the puzzle of defining language. This was because so many of them were uneasy with that Whorfian construct, the 'thought world'. For his ideas to acquire substantive content, they believed, those ideas should in principle support inferences from 'language' to 'culture' (or, say, from linguistic results to pertinent ethnographic questions); so then 'language' had better not be coterminous with 'culture', but rather be at once theoretically and empirically extricable from it. No consensus emerged, however, on how best to specify the conditions Whorfian relativity should meet to achieve legitimacy in this way; on what the chances for success were; or, finally, on what salutary lessons for directing future research might be learnt from the Whorfian problematic.

In contrast, my version of 'the guidance and constraint hypothesis' was deliberately phrased to afford some protection from the dilemma, at least initially: the intention was that investigating the relation between fundamental syntactic and semantic features of a language and the philosophy done in that language should remain within 'language' as such, as distinct from 'culture'. Furthermore, if Chinese philosophy is to constitute a unity 'shaped' or 'coloured' by Chinese language, the latter must form a linguistic foundation which can be perceived to vary neither synchronically nor diachronically. Unlike the field anthropologist, therefore, who works on communities inhabiting real stretches of space and time, the philosophical neo-Whorfian will be forced to dive deep below the surface phenomena and stay there, if the thesis of an integrated type of distinctively Chinese thought is to be maintained.

责任编辑：admin