Postscript on Denoting

I received some criticism of my post yesterday on the value of studying indigenous languages, to the effect that I was confusing denotation and connotation in my discussion of the difficulty of translating paradigm sentences used in Anglophone philosophy of language. 

I assure you, I know the difference. It is my argument however that what we pick out as capable of being denoted --cats, mats, floors, vixens-- is determined entirely by what is of interest to us in a connotative way. If this is true, it of course means that there is no hope of ever finding any settled analytic truths (at least as concerns the natural world). “Gold ist ein gelbes Metall,” to take Kant’s classic example, can’t even be translated from German into English in a way that is satisfying to anyone: gold isn’t ‘yellow’. It’s golden! Setting that problem aside, Kant’s definition also entails that a blind --or even just a colourblind-- person could never really know gold. Nor did Kant know about atomic numbers, which I think it is fair to say have rightly displaced superficial features of chemical elements such as colour (in a certain light, to a certain optical system) in getting at the true and proper definition of, for example, gold. 

Such questions have been beaten to death by now, and I’m not here to revive them, exactly. But I do think that Williamson's shift from ‘gold’ to ‘vixen’ as a paradigmatic example of an analytic truth is peculiar in that it moves us onto less solid ground, whereas one would think that in the weakening of antiquated examples of analytic truths drawn from chemistry that comes with the improvement of chemical knowledge, we would do well to not go looking for new examples in a far less stable domain of human knowledge such as zoology. With gold as with vixens, what is going to strike a given person as analytic depends on what they already know (e.g., again, if you do not know post-Mendeleevian chemistry, ‘Gold is atomic number 79’ is not going to strike you as a statement with full equivalency between subject and predicate). But still, gold is something I think we all think we know about. Not so with vixens, which are rather far from our current urban reality (unless perhaps you are in London). Most of us today do not pay attention to the breeding habits of vulpines, and in large part for this reason, I suspect, the majority of occurrences of the word ‘vixen’ in recent English are metaphorical, describing some perceived feature of a human female; and probably not just metaphorical, but also ironic, signalling that while one knows the metaphorical meaning of the term, one is not actually the sort of person who would use it in that way. 

Many words are like this: the use of them has become detached from the thing they were supposed to denote, and is now largely use by extension. Many such words, moreover, are what we may call ‘precious’, and their use is not just metaphorical, but also a sort of display of mastery on the part of the user. ‘Vixen’ is a borderline case, but consider, for example, the (purported) term for a group of crows. I strongly suspect that at least 95% of occurrences of the phrase ‘a murder of crows’ are found in sentences like, “Did you know that a group of crows is called a ‘murder of crows’?” With this in mind, it can’t really be correct to say that a group of crows is a murder; the preponderance of occurrences of the term in sentences of the sort I just gave means that, in the other 5%, the ones where English-speakers say things like, “Look at that murder of crows,” what is in fact happening is that the speaker is drawing attention to the fact that he or she has mastered this precious bit of vocabulary. The focus of the proposition, in other words, is the speaker, and not the crows. 

The more attentive you are to questions of style in language, the more you become attuned to the fact that a great deal of it works in this way. As far as I can tell no one actually uses the word ‘temblor’; it is a hack synonym for journalists to throw in when they have already used the word ‘earthquake’ too many times, and need to show a capacity for variety. I’ve long suspected that people almost never use certain anatomical orthophemisms in a way that faithfully focuses the listener’s attention on the denotation of the term, but instead that whenever we hear ‘penis’ or ‘vagina’, what we’re actually hearing is a sort of performance of the speaker’s maturity. (As I’ve pointed out before, both of these, as well as ‘anus’, have their origins in Latin euphemisms: 'tail', 'sheath', and 'ring', respectively). I never use these words anyway, but always talk around them, aware that they pose an objective and irresolvable problem to anyone who cares about language, and understands that real mastery of language is not just about getting things right, but calibrating one’s expression of what is right so as to allow its performative aspect to be evident only as much as one wishes. A hack by contrast is someone who uses words like ‘temblor’ or ‘penis’, or phrases like ‘to tamp down’, and thinks they’re getting away with it.

In the end I fear there is no clear boundary between these words that seem performative and therefore 'fake’ to me — ‘vixen’, ‘temblor’, ‘murder’ ‘penis’— on the one hand, and the nouns we take to be most solid and certain and uncontroversial on the other — ‘cat’, ‘mat’, ‘gold’, and so on. It is not just that the analytic/synthetic distinction is untenable, as I think Quine decisively showed in the middle of the last century; it is not just that any sentence will potentially contain different information in the predicate than what was contained in the subject in a way that depends on what you already knew; but that before we even get to the predicate, the subjects of our propositions are all, inescapably, charged up with so much strange energy, so much power of implicit revelation about the speaker or writer, as to doom from the start the search for a neutral sample sentence that focuses our minds only upon the clearly and unambiguously denoted objects. “Vixens are female foxes,” or “The cat is on the mat”, anyhow, are not going to serve the purpose monolingual Anglophone philosophers want them to serve, and this becomes painfully clear the minute we try to translate them into other languages, particularly indigenous languages, noticing as we do so all the approximations and half-measures we are compelled to make and take. What might serve then as an example of a sentence whose meaning is uncontroversial? There is no such thing. 

Comments

[Michael Carrithers]

Most excellent. I've read your last two blogs, and find they strike to the heart of what anthropologists of the sociocultural kind have to deal with. We're taking a term or expression in one lifeworld and translating it -- or maybe better, transforming it -- into the smaller space of academic prose, where the performative element, something like 'I, the author, am telling it like it really is!', is pervasive and highly constraining. With any luck, of course, that transformation is also modified, but energized, by the quality of the fieldworker's effort to figure out what the expression meant before transformation.

Then there are two useful strategies to achieve a more or less effective transformation. One is to transliterate the original term (The Buddhist Pali word *dukkha* strikes me as a good example, meaning something in the range between 'unsatisfactoriness and suffering'), and then play with it so as to explore some of the weights the term has in different usages in the original setting. The second is to create a somewhat distorted English term, say by italicizing and English word, or capitalizing it, and then playing with it. Neither of these strategies are at all unique to anthropology, but they have this in common, that they force the attentive reader onto new ground, 'making strange' in order to lead the reader toward increasing familiarity.

In any case the terms 'denotation' and 'connotation' are little help. It's more like a move from no competence to passive competence in another language. Or from complete alienation to partial recognition.

[Tehseen]

One can read this and indulge in self-amusement for hours. That's exactly where I stopped.
"I never use these words anyway, but always talk around them, aware that they pose an objective and irresolvable problem to anyone who cares about language, and understands that real mastery of language is not just about getting things right, but calibrating one’s expression of what is right so as to allow its performative aspect to be evident only as much as one wishes."
What do you think of the work "pound" in context of Iraq, Afghan wars?

[Martin]

Two things:

1) The verb "golden" also exists in German (the non-inflected form is... "golden“). Yet, he did not write "Gold ist ein goldenes Metall". Possible that the German "gelb" is more related to "golden" than "yellow". Arguably, the specific color term would rather be "goldgelb" (i.e. golden yellow), which is an actually existing German word (not a productive composition I just made up), while "golden" might also mean something like "of gold". But then again, maybe not, and the sentence "Gold is a yellow metal" has the same meaning as the German original and thus exactly corresponds. Meaning that the claim that this sentence cannot even be translated needs to be shown, not asserted (German and English are close, after all), just as the contrary assertion of course.

2) That atomic numbers are "getting at the true and proper definition of, for example, gold" as compared to "superficial features" is, I think, seriously wrong - at least in the context of the present example. Kant characterizes it in two ways: It has a certain color, and it is a metal. Both terms refer to *bulk* properties of a connected ensemble of many atoms. The atom itself has none of these properties: An atom has no color, and it also has no metallic properties. To the extent that elements are called metals, it refers to the bulk property of the element at standard conditions mapped back onto the periodic table, i.e. *exactly* the superficial characteristics that Kant uses. That the bulk properties are already set by the atomic characteristics is neither here nor there: It's the bulk, an emerging property, that *makes* the property, not the atom. (A human body is not a list of its organs, neither are those a list of their cells, or these of their organelles, etc.) In other words, talking about the atom and the metal just is not the same thing: The first is a homonym to the second that just was not available to Kant, not a "superficial property":

Homonym 1: gold = yellow, a metal, etc.
Homonym 2: gold = a chemical element with a given number of protons etc.

These are of course related in that the first is made up of the second, but the first, crucially, also has interactions between them: The properties of the first are *not* the same as those of the second, but *emergent* (bulk properties). There is a certain "water is H2O" tradition in semiotics that seems very confused (or rather not sufficiently confused) about this.

[Martin]

@Martin

The *adjective* golden, what the...?!