Words, meaning, and thought 1: heresy and hokum?

In front of me is a relatively newly published book, Words & Meanings, by Cliff Goddard and Anna Wierzbicka (Oxford UP, ISBN 978 0 19 966843 4).

Although it tackles its subject from a fresh angle, that of lexical semantics and a ‘Natural Semantic Metalanguage’ (NSM), it seems to be walking old and hotly contested ground. Over the next few articles in this series, I will discuss its ideas and their impact, and how they relate to the relationship between language and thought.

Linguistic relativity

The relationship between language and thought has exercised a great many great minds, and is discussed very cogently in this Wikipedia article. Many if not most of us have the idea (however irrational) that language influences thought. In the nineteenth century, with its empires and what might be termed empirical ideas, language was seen by some as the expression of the ‘spirit’ of a nation, or culture.

The European nations, with their long, proud written cultural histories and intimately documented languages full of nuance and sophistication, were of course civilised and sophisticated, and hosted the most elevated of thought. We still make associations today, for instance between Classical Greek and Latin and the founding fathers of European civilisation as we like to think of it – not so much an Anglocentric view, more an Indo-European view (and Sanskrit is another rich addition to that case).

Although linguistic relativity has become known as the ‘Sapir-Whorf hypothesis’, it is generally agreed that associating both Edward Sapir and Benjamin Lee Whorf with it is inappropriate. However there are two versions expressed:

  • the strong form, in which language determines thought;
  • the weak form, in which linguistic categories and usage influence thought and some forms of non-linguistic behaviour.

Misreading words

One particularly common manifestation of both of these versions is the claim that, because a particular language lacks a word for something, that reflects on an absence or unimportance in that culture or society. The corollary of course is when it is claimed that a language has many words for a particular something, then that establishes how that culture or society values that something, or has deep understanding of it.

The ‘language X lacks word Y’ meme is a popular topic on the excellent Language Log blog, and even makes its way into figurative speech:
“His manner can be like that of a general: He’s always right and does not know the word sorry.” (Louise Betts Egan, I Remember when Dad… Memories and Stories about Fathers, 2003, Andrews McMeel Publishing, Missouri, p. 69.)

My own favourite examples are the words mittelschmerz and schadenfreude in German, and hygge in Danish.

Mittelschmerz is a single word which describes a characteristic abdominal pain which some women suffer quite regularly at the time of ovulation, in the middle of their menstrual cycle. I am not aware of its literal equivalent in any other language, and the only English equivalent is the long phrase that I have given. There is no evidence that it is confined to German women, or any more common or severe in German women. Tens of millions of English-speaking women suffer from it every cycle, even though they cannot name it in a single word.

Schadenfreude is the derivation of pleasure or satisfaction from the misfortune and suffering of others. I am not aware of its literal equivalent in any other language, and the only English equivalent is the long phrase that I have given. It is an unfortunate quality of many humans of all nations and cultures, and most certainly is not confined to, nor unusually common in, Germans. There are plenty of English-speaking people who sadly suffer from it as badly as anyone else, even though we are unable to brand them with a single word to describe their quality.

Hygge is a warm, cozy, amicable atmosphere and ambience that, if you have ever visited Denmark, you will surely have enjoyed. I am not aware of its literal equivalent in any non-Scandinavian language, and the only rough English equivalent is that rather muzzy impression that I just tried to give. Although often described as being peculiarly Danish, particularly by those promoting tourism in Denmark, it is by no means confined to Danes or Denmark, and I have experienced it in many societies around the world, even in Norway, Sweden, and Finland (although I have avoided mentioning that to my Danish friends).

There are seemingly endless numbers of other words which exist in only one, or a few, languages but which describe common objects and ideas. One area that has been studied in depth is colour naming: for example, until recently when it borrowed the English word, Welsh had no colour term for the colour pink. It would clearly be absurd to suggest that Welsh-speakers have only recently been able to see pink, or to discriminate pinks from related colours.

Indeed the English word pink was until the middle of the nineteenth century used to describe a vaguely brown to yellow or even greenish yellow colour, in the nineteenth century became used for brighter reds even to scarlet, and only became commonly used for light red and flesh tones towards the latter half of that century.

The corollary is best exemplified by the story of the ‘Eskimos’ having many different words for snow. Here this Wikipedia article gives an excellent account, stressing how this myth goes back to Franz Boas, Sapir and Whorf, and how it has been soundly rebutted by Geoffrey Pullum and others.

The lexicon of life

At their very outset, Goddard and Wierzbicka make it clear that they are unmoved by these major problems:

“There is a very close link between the life of a society and the lexicon of the language spoken by it. This applies in equal measure to the outer and inner aspects of life. An obvious example from the material domain is that of food. It is clearly no accident, for example, that Polish has words for cabbage stew (bigos), beetroot soup (barszcz), and plum jam (powidła), which English does not; or that English (the language of the old British Empire) has a word for orange, or orange-like, jam (marmalade), and Japanese a word (sake) for a strong alcoholic drink made from rice. Obviously, such words can tell us something about the eating or drinking habits of the peoples in question.

“What applies to material culture applies also to people’s values, ideals, and attitudes and to their ways of thinking about the world and our life in it.” (op. cit. p 8).

Had Goddard and Wierzbicka been more systematic and inquisitive, and less glib, they might have considered that many items in the staple or stereotypical diet of English-speakers in the UK (who of course represent a tiny fraction of English-speakers worldwide) consist, as in their Polish examples, of two or more words: roast beef (‘of Olde Englande’), bread and butter, chip butty, fish and chips, and so on.

They might also, by looking no further than the OED, have discovered that marmalade, a word acquired from the French for a type of apple grafted on a quince, was originally not made from oranges, but quinces, and has also been used as a term for plum jam, which remains a fairly popular conserve in the UK (as well as Poland, it would appear). As marmalade is recorded as early as 1524, long before the days of the rather younger British Empire, their inference is a modern misconstruction.

In the beginning, the word?

Before anyone can start trying to make comparisons between different languages, they have to overcome the great differences between languages in their handling of words. The most critical might claim that the whole concept of a word is an artifice, given that until very recently most languages were overwhelmingly oral, and in speech separating individual words may not always be easy.

Linguistics is full of issues which made it hard to know exactly what is a word, and how to compare words between languages. There are clitics, for example, which might start off as separate words, and eventually end up as affixes, but for centuries exist somewhere in between those two states.

Languages such as German are well known for agglomerating word fragments into single, large, and (to non-Germans at least) daunting compounds, of which mittelschmerz and schadenfreude are but pale and terse examples. English has tended to keep words in greater isolation, although there have been moments when it has thrown back to more ancestral tendencies, in monstrosities such as antidisestablishmentarianism, beloved by many a schoolboy.

Goddard and Wierzbicka’s quest for the atomic particles of meaning stumbles over this first problem without clearly identifying it.

Indeed, their approach to the importance of words is flexible only when they want it to be. Having proposed a set of ‘lexical primes’, these atomic particles of meaning, they are happy to offer dispensation in order to fit the data to the hypothesis. Considering a claim that Yucatec Maya lacks words corresponding to after and before, they write:

“but he does not deny that relations of temporal sequence can be clearly conveyed in the language.” … “From an NSM point of view, however, the crucial thing is that if a subordinate clause marked with ts’o’k and depicting event A is followed by a main clause depicting event B, then the meaning conveyed corresponds to English ‘After A, B’. The fact that the interpretation is different if ts’o’k appears instead in the main clause (where it will correspond roughly to ‘finish’) suggests that ts’o’k is polysemous.” (op. cit. p. 16).

Words and meanings

Assuming that we could somehow make meaningful comparisons between languages at a word level, the next problem is how we can determine the meaning of any given word in a given language. Goddard and Wierzbicka are correct in asserting:

“the Anglophone linguistic tradition has generally paid little attention to word meanings. In contrast, however, outside linguistics many Anglophone writers, literary critics, philosophers, historians, and others have shown an acute awareness of the importance of words and their meanings in human life and in the life of societies.” (op. cit. p. 19).

Given that said Anglophone or Francophone (or any other -phone which you care to consider) writers, literary critics, philosophers, and historians are using those words in rhetoric and argument, whilst linguists are examining with the inner workings of language, this is perhaps an unsurprising state of affairs.

However Goddard and Wierzbicka reject lexicography as a means of discovering meaning, as they consider dictionary definitions to be circular and untranslatable (op. cit. p 3). Instead they adopt a metalanguage, whose atomic particles, 65 ‘semantic primes’ are deemed “simple universal concepts”, and therefore apparently need no definition (op. cit. p. 11).

To give you an idea of how these work in practice, here are the derived meanings, their term ‘explications’, of the non-prime concepts of women and men (op. cit. p. 37).

a. people of one kind
b. people of this kind are not children [m]
c. people of this kind have bodies of one kind
d. the bodies of people of this kind are like this:
e. inside the body of someone of this kind there can be for some time a living body of a child [m]

a. people of one kind
b. people of this kind are not children [m]
c. people of this kind have bodies of one kind
d. the bodies of people of this kind are not like women’s [m] bodies
e. some parts of bodies of this kind are not like parts of women’s [m] bodies

I suspect that sundry writers, literary critics, philosophers, and historians with their acute awareness of the importance of words and their meanings might be a little surprised.

My next article will go on to consider those semantic primes.