Do languages get (all) the words they need?

Commenter Nick on Mark Liberman’s Language Log posting on Icelandic peculiarities, in particular, the lack of a word equivalent to the English politeness marker please:

there’s a couple of points where English is lacking in comparison with other languages.

… 2. The English language can ask questions of quantity only when the answer is cardinal, not ordinal. The answer is “Barack Obama is the 44th President of the USA.” I know “Barack Obama” and I know “President of the USA”, but I need to know the ‘nth’ part. What question do I ask in English?

What’s wrong with English?

First observation: a number of languages do in fact have a wh-question word awkwardly translated into English as “the how manieth/manyeth”. Both editions of Timothy Shopen’s Language Typology and Syntactic Description (1st, 1985; 2nd, 2007) have brief discussions in their sections on constituent interrogatives (in their chapters on speech act distinctions in grammar in volume I): by Jerrold Sadock and Arnold Zwicky in 1985, by Ekkehard König and Peter Siemund in 2007. Question expressions for ordinal numbers are a minor type, but well attested.

Just sticking to Europe, König & Siemund mention Finnish monesko and German der wievielte. Der wievielte as in Der wievielte ist es heute? ‘What is today’s date?, What day of the month is it?’ and zum wievielten Male? ‘how many times?’ (somewhat adapting examples from a German dictionary, a Wildhagen’s of some vintage, that happened to be at hand).

Second observation: It’s clear from these examples, and from queries on the web, that translating interrogative ordinal expressions from other languages into English is no easy matter. My Wildhagen’s gives the gloss ‘which (of a number)’ for der wievielte, which can be made somewhat clearer and more accurate by expansion, along the lines of ‘which (in a sequence of a number of things)’ — but these are pointers towards the semantics of the expression, not suggestions as to translations (not that anyone should take dictionary glosses off the shelf as translations of the items glossed).

Instead, translations have to be devised for each use afresh, though undoubtedly there are families of translation strategies, taking in sets of uses. These can require ingenuity, and often some circumlocution, especially out of context. (In context, things can be easier. So if I’m told that of five named children in a family, Alice is the second child and Thomas the fourth, I can ask, “Which child is Kim?”, and be understood as asking about Kim’s serial order among the siblings. Without such a context, my question would not be understood that way.)

Third observation: At least some English speakers, some of the time, have seen the need for a short fixed expression to fill the gap, and have innovated — with how manyeth, how manieth, or how manyth (sometimes with a hyphen after the how, often not) — to serve the purpose; excluding explicit discussions of the appropriateness of the innovations, all are attested in informal writing on the net.

These variants are not hard to understand (in context), but they haven’t caught on, and (at least for the moment) don’t look like having much of a future in generally used ordinary English. That is, no matter how much some people think English needs a word of this sort, it seems unlikely that the gap is going to be filled soon.

Fourth observation, on what “word” means in talk about languages having or not having “a word for” some concept. There are more details beyond the three I’m going to mention briefly, but these are the top three, and they’re the ones I treated at greater length in a Language Log posting, here (to which I refer you).

Ordinary language (OL): We’re looking for expressions in ordinary language, not technical language or other specialized language. Pick a concept that some people might have reason to refer to, and someone has probably coined an expression for it, but so long as this expression is confined to use by certain groups in certain contexts, it’s not relevant to this discussion (though specialized languages are of interest in their own right).

Fixed expression (FE): “Word” in these discussions clearly has to take in some expressions larger than a single word, a consideration that comes up again and again over the candidates in the various categories in the American Dialect Society’s Word of the Year competition. On the other hand, expressions that are semantically compositional (like male cousin) are obviously beside the point. So: fixed expressions.

Of some currency (SC): We should also exclude archaic, dialectal, etc. expressions, interesting though they are in their own right.

Putting all the factors together, we’re looking at ordinary-language fixed expressions of some currency, OLFESCs for short. (I owe the convenient acronym to Geoff Nunberg. OLFESC itself is, of course, not an OLFESC).

Fifth observation: A language’s stock of OLFESCs will have some connection to the concerns of the people who speak the language, that is, to the culture the language is embedded in. In particular, there’s not much reason for a language to have an OLFESC for some concept to which the speakers of the language are indifferent, or of which they are ignorant. (So the OLFESCs of other languages can sometimes present grave problems for translation into English.)

The question is how far these things go. Consider the English OLFESC arm pit / armpit, referring to the axilla, the hollow under the arm at the shoulder. English has no OLFESC for the hollow at the elbow or the hollow at the knee. It’s not that English speakers lack the concepts; we have the compositional expressions hollow at the elbow and hollow at the knee, in fact. What we don’t have is OLFESCs, though elbow pit and knee pit have been coined several times, especially by children, and are easily comprehended.

A number of people have suggested to me that since arm pits have specialized sweat glands, which give rise to culturally significant odors, while elbow pits and knee pits produce only ordinary sweat, it makes sense for a language (like English) to have an OLFESC for the first but not for the others. This strikes me as a plausible idea, but not fully persuasive.

In other cases, it seems pretty clear to me that gaps in the stock of OLFESCs have a considerable component of (historical) accident to them. For instance, the absence of sex-marked counterparts to sex-neutral cousin, while otherwise parallel kin concepts have only sex-marked OLFESCs in English (niece/nephew, aunt/uncle), and languages not very distant from English have sex-marked OLFESCs in the cousin domain, doesn’t seem to follow from anything in the culture. (The explanation seems to be historical: English got cousin from French, but lost the gender-marking on cousin along with other morphological marks of gender.)

So I suggest that languages don’t always get all the words (that is, OLFESCs) they need, or even all the words/OLFESCs their speakers might want.

There’s a lot more to be said on the topic, of course.