Search

There is general agreement on the main features of the process through which the phonology of modern standard Chinese has evolved over three millennia from that of Old Chinese. However, according to general linguistic theory, that phonological history is impossible: the theory claims that no human language can evolve in the manner in which Chinese is believed to have evolved. Furthermore, this particular strand of general linguistic theory has recently been corroborated through stringent statistical testing. Thus there is a glaring contradiction between two areas of scholarship, and to date there has been little recognition by the scholarly community of the need to resolve this contradiction, indeed little willingness to admit its existence. I argue that the contradiction is real and serious, and needs resolution.

Whether spelling is phonemic is only one factor in whether a writing system is suitable for a given language; another consideration is "lexical distinctiveness" of spelling, in two separate senses, and the importance of lexical distinctiveness grows as a writing system matures. Seen in this light, the increasing irregularity of English spelling, over the period when English has been a written language, has been a desirable response to the changing needs of the English-speaking world.

Recent cross-language research has yielded strong statistical evidence in support of the idea, advocated by Andre Martinet and widely accepted by linguists, that languages avoid adopting sound-changes which would create many homophones. Yet we know that the history of Chinese phonology has been marked by repeated phoneme mergers and losses which led to a very high incidence of homophony, forcing the monomorphemic vocabulary of the classical language to be replaced by a largely bimorphemic modern vocabulary. This paper examines various ways in which this apparent contradiction might be resolved. None seems fully satisfactory, yet some resolution must exist.

This ‘short shot’ article draws attention to an unresolved contradiction between fundamental assumptions of modern linguistics and economics. Endogenous growth theory has become the consensus explanation of the phenomenon of long-term economic growth. This theory makes an assumption that is not seen as controversial by economists, namely, that the supply of new, unanticipated ideas is limited only by the level of resources committed to idea generation. Much linguistic theorizing, by contrast, assumes that potential human cognitive products are tightly constrained by biology. Although linguists and economists are primarily interested in different kinds of ‘ideas’, there is a real contradiction between the respective assumptions. The article does not offer a resolution of the contradiction, but encourages members of our discipline to be aware of it as a problem needing to be resolved.*

Patrick Hanks sees linguistic approaches to word meaning as divided between two unattractive extremes. Generative theories, such as were pioneered by Katz and Fodor (1963) and pursued recently e.g. by Wierzbicka (1996), attempt to capture meanings with an apparatus of quasi-mathematical rules and universal semantic primitives which is unequal to reflecting the messy realities revealed by empirical corpus studies. On the other hand, the doctrine of linguistic creativity advanced by Sampson (1980, 2001) is unduly defeatist in denying the possibility of scientific analysis. Hanks argues that theoretical linguistics and practical lexicography should both embrace an intermediate position which distinguishes between high-frequency “norms” of usage and rare “exploitations”. This allows linguists and lexicographers to produce scientific lexical description while nevertheless acknowledging messy variability.

Linguistic science of the past half-century has often been distorted through neglect of normal scientific standards of empirical falsifiability. An earlier paper in this journal used a quantitative literature survey to examine how far in practice the newer trend towards use by linguists of corpora and other empirical data sources had progressed. The result was ambiguous: a trend towards greater empiricism had occurred since about 1970, but around the turn of the century it appeared to have reversed, and the end of the period surveyed (2002) fell so soon thereafter that it was hard to guess whether this reversal was a blip or a long-term change. With a further decade of linguistic literature to examine, the present paper repeats the survey using a more systematic sampling technique, and this yields results that are much more clearcut than those of the earlier paper.

Anyone who urges that differences between languages may correlate with differences in societies' perceptions of the world is open to misunderstanding by those who do not recognise the arbitrariness of their own socially-conditioned perceptions. A striking example is the reception of William Gladstone's nineteenth-century analyses of the vocabulary of the Homeric epics, Europe's first literature. Gladstone anticipated themes that are commonly seen as original advances of twentieth-century anthropology and linguistics; but this achievement has been obscured by a longstanding misinterpretation, according to which Gladstone ascribed Homer's surprising use of colour words to colour-blindness. At present, that misinterpretation is being disseminated more widely than ever before. In fact, Gladstone explicitly did not believe that Ancient Greeks were colour-blind. He did express a range of ideas standardly credited to much more recent scholarship. The reception of Gladstone's Homeric writings demonstrates the strength of the human disposition to trivialize significant cultural differences.

In recent decades there has been a trend towards greater use of empirical data, for instance corpus data, within linguistics. I analyse a sample of linguistics articles from the past half-century in order to establish a detailed profile for this trend. Based on consistent criteria for classifying papers as evidence-based, intuition-based, or neutral, the resulting profile shows that the trend (i) is real, but (ii) is strikingly weaker in general linguistics than in the special subfield of computational linguistics, and (iii) appears to have begun to go into reverse.