The short version:
- Place too much faith in a ‘gloss’ for a foreign word and you may well end up looking as foolish as the folk who follow their satnav robotically into oblivion.
An allegory from Google/Apple maps for all users of bilingual dictionaries:
As drivers become enslaved to their satnavs (in the well-mapped world) we increasingly hear ridiculous stories of people driving onto airport runways, into rivers and goodness knows what else. And it’s all the satnav’s fault. As always our digital servants have turned into digital masters and we blindly follow them.
But of course we know it’s not the satnav’s fault. Computer error is just some combination of human errors, either of the programmer or the user. (As a programmer I’m going to blame the idiot user and as a user I’ll blame the idiot programmer.) Users can blindly trust their computers far more than the programmers would ever warrant.
Quite apart from the difficulty of keeping maps up to date, there is a degree of uncertainty inherent in any positioning technique, whether GPS or WiFi triangulation. The more signal from GPS satellites and geolocated WiFi stations the better, but sometimes the uncertainty can be quite large. For that reason sensible applications display this uncertainty in a similar manner in which scientists report ‘error bars’ in charts. This is to avoid the age-old confusion of precision for accuracy.
So my mapping app on my desktop computer may show my best-guess location as somewhere in the street outside my house, but it covers its back by supplying a large circle within which it thinks I’m most likely to be located. In the case of a computer (without GPS) this is fairly conservative, that is, a large circle about a Glasgow street block in diameter.
On my smartphone with the aid of GPS the location is different, not precisely correct, but closer and the uncertainty is much smaller. However, there’s still a blue circle of uncertainty. If I evaluate the location based on the centre pin I would say both the GPS-assisted phone and the computer miss the mark, but in both cases I’m within the blue circle of uncertainty (or is it certainty?). The proper way to use these tools is to recognise that I am pretty much guaranteed to be somewhere within the domain of the blue circle, but there’s no guarantee that I’m actually in the middle of it. The more perspectives and sources of information the better. The centre pin quickly gets us to close to the location, but we shouldn’t assume it is precisely correct. Now, why am I labouring so much that is obvious?
It’s because in the world of translation, especially instant digital translation, it is very easy to fall into the trap of mistaking precision for accuracy, and especially where ‘glosses’ are concerned. Glosses are the short equivalents we give for foreign words, such as the gloss ‘hand’ for Greek χειρ, Hebrew יַד and Hausa hannu. The gloss ‘hand’ is like the centre pin in that map; they use our knowledge of English to get us very quickly somewhere close to what a Greek/Hebrew/Hausa speaker may mean by those terms. But! (And there is a big but.) This does not mean that Greek χειρ actually means the same thing as ‘hand’ any more than the centre pin shows exactly where I am.
- χειρ seems to include the body from the fingertips past the wrist and perhaps up to the elbow. (But it’s hard to get that out of standard dictionaries, since ‘hand’ seems so obvious.)
- יָד includes fingers, hands/palms, wrists and forearms, like Greek. (This is more clearly stated in NIDOTTE*.) כַּף is the palm, so that it might actually be more helpful to gloss יָד as ‘arm’ and כַּף as ‘hand’. The word may have become more general over time.
- hannu is the whole upper limb (arm+hand). This seems to be more than the Greek and Hebrew terms typically encompass.
That’s what the words mean, but they also are used in a variety of ways. Sometimes that would result in the word ‘hand’ being used in an equivalent expression in English and other times not. (For example ‘παραδιδόναι τινὰ εἰς χεῖράς τινος’ might be glossed ‘give something into hand of someone’ but more properly translated into English as ‘hand something over to someone’ where the word ‘hand’ actually is used as a verb in a way that Greek cannot achieve.)
Let’s return to the GPS allegory. The centre location may be good enough for a cruise missile to do its work but not good enough for a sniper. If we’re a sniper (or maybe and more pleasantly an Amazon delivery drone operator) then we would need to open our eyes and see how closely the approximation does match up to the reality for our purposes, taking into account that the real location could be anywhere within the blue circle.
Why do we use glosses? They are a short-cut. They help us leverage our existing knowledge to get a good handle on what someone else means. But we haven’t necessarily ‘reached out final destination’ until we open our eyes and understand the differences that there are between the languages.
How do we open our eyes?
- We can observe how the words are actually used in phrases and idioms.
- We need to look out for telltale signs that our gloss doesn’t quite match up right.
(For example Jacob crosses his ‘hands’ in Gen. 48:14 clearly meaning ‘forearms’; or someone may indicate a pain in their ‘hand’ while pointing to their upper arm. My wife being asked by a Nigerian whether our 1-year-old was ‘a baby or a man’; the English words ‘baby’ and ‘man’ were inadequate glosses for underlying Hausa terms used to distinguish girls and boys.)
- We may need to ask people dumb questions about what they really mean. Of course some times it really doesn’t matter because the level of precision required isn’t so great.
- Remember that some words cover a general area of meaning just as the location ‘Glasgow’ would properly be a surface or area rather than a point. Generality and specificity will vary from one language to another, as will the ordinariness of terms.
This is all what makes translation such a fun and infuriating business. And it’s also why I’m writing a paper on making dictionaries serve translation.